There has been over the past few years, a very increased popularity for yoga. A lot of literatures have been published that claim yoga to be beneficial in improving the overall lifestyle and health especially in rehabilitation, mental health and more. Considering the fast-paced lives that individuals live, people usually prefer to exercise or work-out from the comfort of their homes and with that a need for an instructor arises. Hence why, we have developed a self-assisted system which can be used to detect and classify yoga asanas, which is discussed in-depth in this paper. Especially now when the pandemic has taken over the world, it is not feasible to attend physical classes or have an instructor over. Using the technology of Computer Vision, a computer-assisted system such as the one discussed, comes in very handy. The technologies such as ml5.js, PoseNet and Neural Networks are made use for the human pose estimation and classification. The proposed system uses the above-mentioned technologies to take in a real-time video input and analyze the pose of an individual, and classifies the poses into yoga asanas. It also displays the name of the yoga asana that is detected along with the confidence score.
Efficient and accurate object detection has been an important topic in the advancement of computer vision systems.
Our project aims to detect the object with the goal of achieving high accuracy with a real-time performance.
In this project, we use a completely deep learning based approach to solve the problem of object detection.
The input to the system will be a real time image, and the output will be a bounding box corresponding to all the objects in the image, along with the class of object in each box.
Objective -
Develop a application that detects an object and it can be used for vehicles counting, when the object is a vehicle such as a bicycle or car, it can count how many vehicles have passed from a particular area or road and it can recognize human activity too.
Slides by Amaia Salvador at the UPC Computer Vision Reading Group.
Source document on GDocs with clickable links:
https://docs.google.com/presentation/d/1jDTyKTNfZBfMl8OHANZJaYxsXTqGCHMVeMeBe5o1EL0/edit?usp=sharing
Based on the original work:
Ren, Shaoqing, Kaiming He, Ross Girshick, and Jian Sun. "Faster R-CNN: Towards real-time object detection with region proposal networks." In Advances in Neural Information Processing Systems, pp. 91-99. 2015.
Efficient and accurate object detection has been an important topic in the advancement of computer vision systems.
Our project aims to detect the object with the goal of achieving high accuracy with a real-time performance.
In this project, we use a completely deep learning based approach to solve the problem of object detection.
The input to the system will be a real time image, and the output will be a bounding box corresponding to all the objects in the image, along with the class of object in each box.
Objective -
Develop a application that detects an object and it can be used for vehicles counting, when the object is a vehicle such as a bicycle or car, it can count how many vehicles have passed from a particular area or road and it can recognize human activity too.
Slides by Amaia Salvador at the UPC Computer Vision Reading Group.
Source document on GDocs with clickable links:
https://docs.google.com/presentation/d/1jDTyKTNfZBfMl8OHANZJaYxsXTqGCHMVeMeBe5o1EL0/edit?usp=sharing
Based on the original work:
Ren, Shaoqing, Kaiming He, Ross Girshick, and Jian Sun. "Faster R-CNN: Towards real-time object detection with region proposal networks." In Advances in Neural Information Processing Systems, pp. 91-99. 2015.
Object detection is a computer technology related to computer vision and image processing that deals with detecting instances of semantic objects of a certain class (such as humans, buildings, or cars) in digital images and videos. Well-researched domains of object detection include face detection and pedestrian detection. Object detection has applications in many areas of computer vision, including image retrieval and video surveillance.
human activity recognization using machine learning with data analysisVenkat Projects
Human activity recognition, or HAR for short, is a broad field of study concerned with identifying the specific movement or action of a person based on sensor data.
The sensor data may be remotely recorded, such as video, radar, or other wireless methods. It contains data generated from accelerometer, gyroscope and other sensors of Smart phone to train supervised predictive models using machine learning techniques like SVM , Random forest and decision tree to generate a model. Which can be used to predict the kind of movement being carried out by the person, which is divided into six categories walking, walking upstairs, walking down-stairs, sitting, standing and laying?
MLM and SVM achieved accuracy of more than 99.2% in the original data set and 98.1% using new feature selection method. Results show that the proposed feature selection approach is a promising alternative to activity recognition on smart phones.
A comprehensive tutorial on Convolutional Neural Networks (CNN) which talks about the motivation behind CNNs and Deep Learning in general, followed by a description of the various components involved in a typical CNN layer. It explains the theory involved with the different variants used in practice and also, gives a big picture of the whole network by putting everything together.
Next, there's a discussion of the various state-of-the-art frameworks being used to implement CNNs to tackle real-world classification and regression problems.
Finally, the implementation of the CNNs is demonstrated by implementing the paper 'Age ang Gender Classification Using Convolutional Neural Networks' by Hassner (2015).
Final Year Project BCA Presentation on Pic-O-SticaSharath Raj
This slide is based on the final year project of BCA. Project was on Online image purchase and Sales System.
The system was developed using PHP at the frontend and Mysql at the Backend.
Image will be uploaded and will be watermarked. USer can buy or sell their lovely images.
Object detection is a computer technology related to computer vision and image processing that deals with detecting instances of semantic objects of a certain class (such as humans, buildings, or cars) in digital images and videos.
A Small Helping Hand from me to my Engineering collegues and my other friends in need of Object Detection
Object detection is a computer technology related to computer vision and image processing that deals with detecting instances of semantic objects of a certain class (such as humans, buildings, or cars) in digital images and videos. Well-researched domains of object detection include face detection and pedestrian detection. Object detection has applications in many areas of computer vision, including image retrieval and video surveillance.
human activity recognization using machine learning with data analysisVenkat Projects
Human activity recognition, or HAR for short, is a broad field of study concerned with identifying the specific movement or action of a person based on sensor data.
The sensor data may be remotely recorded, such as video, radar, or other wireless methods. It contains data generated from accelerometer, gyroscope and other sensors of Smart phone to train supervised predictive models using machine learning techniques like SVM , Random forest and decision tree to generate a model. Which can be used to predict the kind of movement being carried out by the person, which is divided into six categories walking, walking upstairs, walking down-stairs, sitting, standing and laying?
MLM and SVM achieved accuracy of more than 99.2% in the original data set and 98.1% using new feature selection method. Results show that the proposed feature selection approach is a promising alternative to activity recognition on smart phones.
A comprehensive tutorial on Convolutional Neural Networks (CNN) which talks about the motivation behind CNNs and Deep Learning in general, followed by a description of the various components involved in a typical CNN layer. It explains the theory involved with the different variants used in practice and also, gives a big picture of the whole network by putting everything together.
Next, there's a discussion of the various state-of-the-art frameworks being used to implement CNNs to tackle real-world classification and regression problems.
Finally, the implementation of the CNNs is demonstrated by implementing the paper 'Age ang Gender Classification Using Convolutional Neural Networks' by Hassner (2015).
Final Year Project BCA Presentation on Pic-O-SticaSharath Raj
This slide is based on the final year project of BCA. Project was on Online image purchase and Sales System.
The system was developed using PHP at the frontend and Mysql at the Backend.
Image will be uploaded and will be watermarked. USer can buy or sell their lovely images.
Object detection is a computer technology related to computer vision and image processing that deals with detecting instances of semantic objects of a certain class (such as humans, buildings, or cars) in digital images and videos.
A Small Helping Hand from me to my Engineering collegues and my other friends in need of Object Detection
Yoga pose detection using deep learning project PPT.pptxssuser4f92fb
idea behind this yoga pose detection project using deep learning or neural network learning is that yoga popularity is increasing day by day because of its benefits. Doing yoga helps us physically, mentally as well as spiritually. Because of this many people nowadays are doing it regularly. The main idea of this project is to help the people to recognize which yoga pose they are doing with the help of this detection technique. Yoga which involves 8 rungs and limbs of it, which includes Yama, Niyama, Asana, Pranayama, Dharana, Dhyana and Samadhi. To easily help people understand which pose they are performing via images, video recording by classifying it, we are implementing this project because of this people will incline towards doing more as they will get help to identify which pose they are doing very easily.
Analysis of student sentiment during video class with multi-layer deep learni...IJECEIAES
The modern education system is an essential part of the rise of technology. The E-learning education system is not just an experimental system; it is a vital learning system for the whole world over the last few months. In our research, we have developed our learning method in a more effective and modern way for students and teachers. For significant implementation, we are implementing convolutions neural networks and advanced data classifiers. The expression and mood analysis of a student during the online class is the main focus of our study. For output measure, we divide the final output result as attentive, inattentive, understand, and neutral. Showing the output in real-time online class and for sensory analysis, we have used support vector machine (SVM) and OpenCV. The level of 5*4 neural network is created for this work. An advanced learning medium is proposed through our study. Teachers can monitor the live class and different feelings of a student during the class period through this system.
Automatic Selection of Open Source Multimedia Softwares Using Error Back-Prop...IJERA Editor
Open source opens a new era to provide license of the software for the user at free of cost which is advantage over paid licensed software. In Multimedia applications there are many versions of software are available and there is a problem for the user to select compatible software for their own system. Most of the time while surfing for software a huge list of software opens in response. The selection of particular software which is pretty suitable for the system from a real big list is the biggest challenge that is faced by the users. This work has been done that focuses on the existing open source software that are widely used and to design an automatic system for selection of particular open source software according to the compatibility of users own system. In this work, error back-propagation based neural network is designed in MATLAB for automatic selection of open source software. The system provides the open source software name after taking the information from user. Regression coefficient of 0.93877 is obtained and the results shown are up to the mark and can be utilized for the fast and effective software search.
Semantic-based visual emotion recognition in videos-a transfer learning appr...IJECEIAES
Automatic emotion recognition is active research in analyzing human’s emotional state over the past decades. It is still a challenging task in computer vision and artificial intelligence due to its high intra-class variation. The main advantage of emotion recognition is that a person’s emotion can be recognized even if he is extreme away from the surveillance monitoring since the camera is far away from the human; it is challenging to identify the emotion with facial expression alone. This scenario works better by adding visual body clues (facial actions, hand posture, body gestures). The body posture can powerfully convey the emotional state of a person in this scenario. This paper analyses the frontal view of human body movements, visual expressions, and body gestures to identify the various emotions. Initially, we extract the motion information of the body gesture using dense optical flow models. Later the high-level motion feature frames are transferred to the pre-trained convolutional neural network (CNN) models to recognize the 17 various emotions in Geneva multimodal emotion portrayals (GEMEP) dataset. In the experimental results, AlexNet exhibits the architecture's effectiveness with an overall accuracy rate of 96.63% for the GEMEP dataset is better than raw frames and 94% for visual geometry group-19 VGG-19, and 93.35% for VGG-16 respectively. This shows that the dense optical flow method performs well using transfer learning for recognizing emotions.
Wearable sensor-based human activity recognition with ensemble learning: a co...IJECEIAES
The spectacular growth of wearable sensors has provided a key contribution to the field of human activity recognition. Due to its effective and versatile usage and application in various fields such as smart homes and medical areas, human activity recognition has always been an appealing research topic in artificial intelligence. From this perspective, there are a lot of existing works that make use of accelerometer and gyroscope sensor data for recognizing human activities. This paper presents a comparative study of ensemble learning methods for human activity recognition. The methods include random forest, adaptive boosting, gradient boosting, extreme gradient boosting, and light gradient boosting machine (LightGBM). Among the ensemble learning methods in comparison, light gradient boosting machine and random forest demonstrate the best performance. The experimental results revealed that light gradient boosting machine yields the highest accuracy of 94.50% on UCI-HAR dataset and 100% on single accelerometer dataset while random forest records the highest accuracy of 93.41% on motion sense dataset.
A Framework for Human Action Detection via Extraction of Multimodal FeaturesCSCJournals
This work discusses the application of an Artificial Intelligence technique called data extraction and a process-based ontology in constructing experimental qualitative models for video retrieval and detection. We present a framework architecture that uses multimodality features as the knowledge representation scheme to model the behaviors of a number of human actions in the video scenes. The main focus of this paper placed on the design of two main components (model classifier and inference engine) for a tool abbreviated as VASD (Video Action Scene Detector) for retrieving and detecting human actions from video scenes. The discussion starts by presenting the workflow of the retrieving and detection process and the automated model classifier construction logic. We then move on to demonstrate how the constructed classifiers can be used with multimodality features for detecting human actions. Finally, behavioral explanation manifestation is discussed. The simulator is implemented in bilingual; Math Lab and C++ are at the backend supplying data and theories while Java handles all front-end GUI and action pattern updating. To compare the usefulness of the proposed framework, several experiments were conducted and the results were obtained by using visual features only (77.89% for precision; 72.10% for recall), audio features only (62.52% for precision; 48.93% for recall) and combined audiovisual (90.35% for precision; 90.65% for recall).
A novel enhanced algorithm for efficient human trackingIJICTJOURNAL
Tracking moving objects has been an issue in recent years of computer vision and image processing and human tracking makes it a more significant challenge. This category has various aspects and wide applications, such as autonomous deriving, human-robot interactions, and human movement analysis. One of the issues that have always made tracking algorithms difficult is their interaction with goal recognition methods, the mutable appearance of variable aims, and simultaneous tracking of multiple goals. In this paper, a method with high efficiency and higher accuracy was compared to the previous methods for tracking just objects using imaging with the fixed camera is introduced. The proposed algorithm operates in four steps in such a way as to identify a fixed background and remove noise from that. This background is used to subtract from movable objects. After that, while the image is being filtered, the shadows and noises of the filmed image are removed, and finally, using the bubble routing method, the mobile object will be separated and tracked. Experimental results indicated that the proposed model for detecting and tracking mobile objects works well and can improve the motion and trajectory estimation of objects in terms of speed and accuracy to a desirable level up to in terms of accuracy compared with previous methods.
Similar to Yoga Posture Classification using Computer Vision (20)
Total Ionization Cross Sections due to Electron Impact of Ammonia from Thresh...Dr. Amarjeet Singh
In the present paper, we have employed modified Khare-BEB method [Atoms, (2019)] to evaluate total ionization cross sections by the electron impact for ammonia in energy range from the ionization threshold to 10 MeV. The theoretical ionization cross sections have been compared to the available previous theoretical and experimental results. The collision parameters dipole matrix squared M_j^2 and CRP also have been calculated. The present calculations were found in remarkable agreement with the available experimental results.
A Case Study on Small Town Big Player – Enjay IT Solutions Ltd., BhiladDr. Amarjeet Singh
Adequately trained Manpower is a problem that affects the IT industry as a whole, but it is particularly acute for Enjay IT Solution. Enjay's location in a semi-urban or rural area makes it even more difficult to find a talented employee with the right skills. As the competition for skilled workers grows, it becomes more difficult to attract and keep those workers who have the requisite training and experience.
Effect of Biopesticide from the Stems of Gossypium Arboreum on Pink Bollworm ...Dr. Amarjeet Singh
Pink bollworm and Lepidoptera development quickly in numbers which is a typical animal group that produces around 100 youthful ones inside certain days or weeks. This assault influences the harvests broadly in the tropical and sub-tropical temperature areas. Thus, to keep up with the yield of harvests the vermin ought to be kept away by utilizing pesticides. The unnecessary measure of the purpose of pesticides influences the dirt, land, and as well as human well-being, and contaminates the climate. Thus, an ozone-accommodating biopesticide is extracted from the stems of the Gossypium arboreum. Thus, the extraction of biopesticide from the stems of Gossypium arboreum demonstrated that the quantity of pink bollworm and Lepidoptera is diminished step by step in the wake of showering the arrangement on the impacted region of the plant because of the presence of the gossypol.
Artificial Intelligence Techniques in E-Commerce: The Possibility of Exploiti...Dr. Amarjeet Singh
E-Commerce has transformed business as we know over the past few decades. The rapid increasing use of the Internet and the strong purchasing power in Saudi Arabia have had a strong impact on the evolution of E-Commerce in the country. Saudi Arabia is yet another country that will release artificial intelligence power to fuel its growth in the economic world. Recently, artificial intelligence (AI) applications that can facilitate e-commerce processes have been widely used. The impact of using artificial intelligence (AI) concepts and techniques on the efficiency of e-commerce, particularly has been overlooked by many prior studies. In this paper, a literature review was conducted to explore and investigate possible applications of AI in E-Commerce that can help Saudi Arabian businesses.
Factors Influencing Ownership Pattern and its Impact on Corporate Performance...Dr. Amarjeet Singh
This study on factors influencing Ownership pattern and its impact on corporate performance has used five industries data viz Automobile industry, IT industry, Banking industry, Oil & Gas industry and pharmaceutical industry for five years from 2017 to 2021. First the factors influencing ownership pattern was identified and later its impact on corporate performance was analysed. Multiple Regression, ANOVA and Correlation was used in SPSS 28. Percentage of independent directors on the board and size of the company has significant impact on Indian Promotor holding and non-institutional ownership has significant impact on corporate performance.
An Analytical Study on Ratios Influencing Profitability of Selected Indian Au...Dr. Amarjeet Singh
Every country with a well-developed transportation network has a well-developed economy. The automobile industry is a critical engine of the nation's economic development. The automobile industry has significant backward and forward links with every area of the economy, as well as a strong and progressive multiplier impact. The automotive industry and the auto component industry are both included in the vehicle industry. It includes passenger waggons, light, medium, and heavy commercial vehicles, as well as multi-utility vehicles such as jeeps, three-wheelers, military vehicles, motorcycles, tractors, and auto-components such as engine parts, batteries, drive transmission parts, electrical, suspension and chassis parts, and body and other parts. In the last several years, India's automobile sector has seen incredible growth in sales, production, innovation, and exports. India's car industry has emerged as one of the best in the world, and the auto-ancillary sector is poised to assist the vehicle sector's expansion. Vehicle manufacturers and auto-parts manufacturers account for a significant component of global motorised manufacturing. Vehicle manufacturers from across the world are keeping a close eye on the Indian auto sector in order to assess future demand and establish India as a global manufacturing base. The current research focuses on three automotive behemoths: TATA Motors, MRF, and Mahindra & Mahindra.
A Study on Factors Influencing the Financial Performance Analysis Selected Pr...Dr. Amarjeet Singh
The growth of a country's banking sector has a significant impact on its economic development. The banking sector plays a critical role in determining a country's economic future. A well-planned, structured, efficient, and viable banking system is an essential component of an economy's economic and social infrastructure. In modern society, a strong banking system is required because it meets the financial needs of the modern society. In a country's economy, the banking system plays a crucial role. Because it connects surplus and deficit economic agents, the bank is the most important financial intermediary in the economy. The banking system is regarded as the economy's lifeline. It meets the financial needs of commerce, industry, and agriculture. As a result, the country's development and the banking system are intertwined. They are critical in the mobilisation of savings and the distribution of credit to various sectors of the economy. India's private sector banks play a critical role in the country's economic development. So The financial performance of private sector banks must be evaluated carefully.
An Empirical Analysis of Financial Performance of Selected Oil Exploration an...Dr. Amarjeet Singh
After the United States, China, and Japan, India was the world's fourth biggest consumer of oil and petroleum products. The nation is significantly reliant on crude oil imports, the majority of which come from the Middle East. The Indian oil and gas business is one of the country's six main sectors, with important forward links to the rest of the economy. More than two-thirds of the country's overall primary energy demands are met by the oil and gas industry. The industry has played a key role in placing India on the global map. India is now the world's sixth biggest crude oil user and ninth largest crude oil importer. In addition, the country's portion of the worldwide refining market is growing. India's refining industry is now the world's sixth biggest. With plans for Reliance Petroleum Limited to commission another refinery with a capacity of 29 MTPA next 16 to its 33 MTPA refinery in Jamnagar, Gujarat, this position is projected to be enhanced. As a consequence, the Reliance refinery would be the biggest single-site refinery in the world. Based on secondary data gathered from CMIE, the current research examines the ratios influencing the profitability of selected oil exploration and production businesses in India during a 10-year period.
Since 1991, thanks to economic policy liberalization, the Indian economy has entered an era in which Indian businesses can no longer disregard global markets. Prior to the 1990s, the prices of a variety of commodities, metals, and other assets were carefully regulated. Others, which were not rolled, were primarily dependant on regulated input costs. As a result, there was no uncertainty and, as a result, no price fluctuations. However, in 1991, when the process of deregulation began, the prices of most items were deregulated. It has also resulted in the exchange being partially deregulated, easing trade restrictions, lowering interest rates, and making significant advancements in foreign institutional investors' access to the capital markets, as well as establishing market-based government securities pricing, among other things. Furthermore, portfolio and securities price volatility and instability were influenced by market-determined exchange rates and interest rates. As a result, hedging strategies employing a variety of derivatives were exposed to a variety of risks. The Indian capital market will be examined in this study, with a focus on derivatives.
Theoretical Estimation of CO2 Compression and Transport Costs for an hypothet...Dr. Amarjeet Singh
SEI S.p.a. presented a project to build a 1320 MW coal-fired power plant in Saline Joniche, on the Southern tip of Calabria Region, Italy, in 2008. A gross early evaluation about the possibility to add CCS (CO2 Capture & Storage) was performed too. The project generated widespread opposition among environmental associations, citizens and local institutions in that period, against the coal use to produce energy, as a consequence of its GHG clima-alterating impact. Moreover the CCS (also named Carbon Capture & Storage or more recently CCUS: Carbon Capture-Usage-Storage) technology was at that time still an unknown and “mysterious” solution for the GHG avoiding to the atmosphere. The present study concerns the sizing of the compression and transportation system of the CCS section, included in the project presented at the time by SEI Spa; the sizing of the compression station and the pipeline connecting the plant to the possible Fosca01 offshore injection site previously studied as a possible storage solution, as part of a coarse screening of CO2 storage sites in the Calabria Region. This study takes into account the costs of construction, operation and maintenance (O&M) of both the compression plant and the sound pipeline, considering the gross static storage capacity of the Fosca01 reservoir as a whole as previously evaluated.
Analytical Mechanics of Magnetic Particles Suspended in Magnetorheological FluidDr. Amarjeet Singh
In this paper, the behavior of MR particles has been systematically investigated within the scope of analytical mechanics. . A magnetorheological fluid belongs to a class of smart materials. In magnetorheological fluids, the motion of magnetic particles is controlled by the action of internal and external forces. This paper presents analytical mechanics for the interaction of system of particles in MR fluid. In this paper, basic principles of Analytical Mechanics are utilized for the construction of equations.
Techno-Economic Aspects of Solid Food Wastes into Bio-ManureDr. Amarjeet Singh
Solid waste is health hazard and cause damage to the environment due to improper handling. Solid waste comprises of Industrial Waste (IW), Hazardous Waste (HW), Municipal Solid Waste (MSW), Electronic waste (E-waste), Bio-Medical Waste (BMW) which depend on their supply & characteristics. Food waste or Bio-waste composting and its role in sustainable development is explained in food waste is a growing area of concern with many costs to our community in terms of waste collection, disposal and greenhouse gases. When rotting food ends up in landfill it turns into methane, a greenhouse gas that is particularly damaging to the environment. Composting is biochemical process in which organic materials are biologically degraded, resulting in the production of organic by products and energy in the form of heat. Heat is trapped within the composting mass, leading to the phenomenon of self-heating. This overall process provide us Bio-Manure.
Crypto-Currencies: Can Investors Rely on them as Investment Avenue?Dr. Amarjeet Singh
The purpose of this study is to examine investors’ perceptions about investing in crypto-currencies. We think that investors trust in crypto-currencies is largely driven by crypto-currency comprehension, trust in government, and transaction speed. This is the first study to examine crypto-currencies from the investor’s perspective. Following that, we discover important antecedents of crypto-currency confidence. Second, we look at the government's role in crypto-currencies. The importance of this study is: first, crypto-currencies have the potential to disrupt the current economic system as the debate is all about impact of decentralization of transactions; thus, further research into how it affects investors trust is essential; and second, access to crypto-currencies. Finally, if Fin-Tech companies or banks want to enter the bitcoin industry may not attract huge advertising costs as well as marketing to soothe clients' concerns about investing in various digital currencies The research sheds light on indecisiveness in the context of marketing aspects adopted by demonstrating investors are aware about the crypto.
Awareness of Disaster Risk Reduction (DRR) among Student of the Catanduanes S...Dr. Amarjeet Singh
The Island Province of Catanduanes is prone to all types of natural hazards that includes torrential and heavy rains, strong winds and surge, flooding and landslide or slope failures as a result of its geographical location and topography. RA 10121 mandates local DRRM bodies to “encourage community, specifically the youth, participation in disaster risk reduction and management activities, such as organizing quick response groups, particularly in identified disaster-prone areas, as well as the inclusion of disaster risk reduction and management programs as part of youth programs and projects. The study aims to determine the awareness to disaster of the student of the Catanduanes State University. The disaster-based questionnaire was prepared and distributed among 636 students selected randomly from different Colleges and Laboratory Schools in the University
The Catanduanes State University students understood some disaster-related concepts and ideas, but uncertain on issues on preparedness, adaptation, and awareness on the risks inflicted by these natural hazards. Low perception on disaster risks are evidently observed among students. The responses of the students could be based on the efficiency and impact of the integration of DRR education in the senior high school curriculum. Specifically, integration of the concepts about the hazards, hazard maps, disaster preparedness, awareness, mitigation, prevention, adaptation, and resiliency in the science curriculum possibly affect the knowledge and understanding of students on DRR. Preparedness drills and other forms of capacity building must be done to improve awareness of the student towards DRRM.
The study further recommends that teachers and instructor must also be capacitated in handling disaster as they are the prime movers in the implementation of the DRRM in education. Preparedness drills and other forms of capacity building must be done to improve awareness of the student towards DRRM. Core subjects in Earth Sciences must be reinforced with geologic hazards. Learning competencies must also be focused on hazard identification and mapping, and coping with different geologic disaster.
The 1857 war was a watershed moment in the history of the Indian subcontinent. The battle has sparked academic debate among historians and sociologists all around the world. Despite the fact that it has been more than 150 years, this battle continues to pique the interest of historians. The war's causes and events that occurred throughout the conflict, persons who backed the British and anti-British fighters, and the results and ramifications, are all aspects of this conflict. In terms of outcomes, many academics believe that the war was a failure for those who started it. It is often assumed that the Indians who battled the British in this conflict were unable to achieve their goals. Many gains accrued to Indians as a result of the conflict, but these achievements are overshadowed by the dispute over the war's failure. This research effort focuses on the war's achievements for India, and the significance of those achievements.
Haryana's Honour Killings: A Social and Legal Point of ViewDr. Amarjeet Singh
Life is unpredictably unpredictable. Nobody knows what will happen in the next minute of their lives. In this circumstance, every human being has the right and desire to conduct their lives according to their own desires. No one should be forced to live a life solely for the benefit and reputation of others. Honour killing is defined as the assassination of a person, whether male or female, who refuses to accept the family's arranged marriage or decides to move her or his marital life according to her or his wishes solely because it jeopardizes the family's honour. The family's supreme authority looks after the family's name but neglects to consider the love and affection shared among family members. I have discussed honour killing in India in my research work. This sort of murder occurs as a result of particular triggers, which are also examined in relation to the role of the law in honour killing. No one can be released free if they break the law, and in this case, it is a felony that violates various regulations designed to safeguard citizens. This crime is similar to many others, but it is distinct enough to be differentiated in the report. When the husband is of low social standing, it lowers the position and caste of the female family, prompting the male family members to murder the girl. But they forget that the girl is their kid and that while rank may be attained, a girl's life can never be replaced, and that caste is less valuable than the girl's life and love spent with them.
Optimization of Digital-Based MSME E-Commerce: Challenges and Opportunities i...Dr. Amarjeet Singh
The impact caused by the Covid-19 Pandemic on Micro and Small and Medium Enterprises (MSMEs) was so severe and fatal
that not a few went out of business. The heavy burden is borne by MSME actors due to social restrictions imposed by the
government, the declining purchasing power of the people, a product that continues to decline until capital runs out. Plus
inadequate knowledge in carrying out marketing strategies and product innovations are the main trigger for the lack of
enthusiasm for MSME actors as well as bankruptcy. MSME digitalization-based e-commerce is an opportunity and the right
solution in dealing with the obstacles caused by the impact of Covid-19, as well as a challenge for MSME actors to design old
ways in new ways through digital business.
Modal Space Controller for Hydraulically Driven Six Degree of Freedom Paralle...Dr. Amarjeet Singh
This paper presents the Modal space decoupled control for a hydraulically driven parallel mechanism has been presented. The approach is based on singular values decomposition to the properties of joint-space inverse mass matrix, and mapping of the control and feedback variables from the joint space to the decoupling modal space. The method transformed highly coupled six-input six-output dynamics into six independent single-input single-output (SISO) 1 DOF hydraulically driven mechanical systems. The novelty in this method is that the signals including control errors, control outputs and pressure feedbacks are transformed into decoupled modal space and also the proportional gains and dynamic pressure feedback are tuned in modal space. The results indicate that the conventional controller can only attenuate the resonance peaks of the lower eigenfrequencies of six rigid modes properly, and the peaking points of other relative higher eigenfrequencies are over damped, The further results show that it is very effective to design and tune the system in modal space and that the bandwidth increased substantially except surge (x) and sway (y) motions, each degree of freedom can be almost tuned independently and their bandwidths can be increased near to the undamped eigenfrequencies.
It is a known fact that a large number of Steel Industry Expansion projects in India have been delayed due to regulatory clearances, environmental issues and problems pertaining to land acquisition. Also, there are challenges in the tendering phase that affect viability of projects thus delaying implementation, construction phase is beset with over-runs and disputes and last but not the least; provider skills are weak all across the value chain. Given the critical role of Steel Sector in ensuring a sustained growth trajectory for India, it is imperative that we identify the core issues affecting completion of infrastructure projects in India and chalk out initiatives that need to be acted upon in short term as well as long term.
A blockchain is a decentralised database that is shared across computer network nodes. A blockchain acts as a database, storing information in a digital format. The study primarily aims to explore how in the future, block chain technology will alter several areas of the Indian economy. The current study aims to obtain a deeper understanding of blockchain technology's idea and implementation in India, as well as the technology's potential as a disruptive financial technological innovation.
Secondary sources such as reports, journals, papers, and websites were used to compile all the data. Current and relevant information were utilised to help understand the research goals. All the information is rationally organised to fulfil the objectives. The current research focuses on recommendations for enhancing India's Blockchain ecosystem so that it may become one of the best in the world at utilising this new technology.
Explore the innovative world of trenchless pipe repair with our comprehensive guide, "The Benefits and Techniques of Trenchless Pipe Repair." This document delves into the modern methods of repairing underground pipes without the need for extensive excavation, highlighting the numerous advantages and the latest techniques used in the industry.
Learn about the cost savings, reduced environmental impact, and minimal disruption associated with trenchless technology. Discover detailed explanations of popular techniques such as pipe bursting, cured-in-place pipe (CIPP) lining, and directional drilling. Understand how these methods can be applied to various types of infrastructure, from residential plumbing to large-scale municipal systems.
Ideal for homeowners, contractors, engineers, and anyone interested in modern plumbing solutions, this guide provides valuable insights into why trenchless pipe repair is becoming the preferred choice for pipe rehabilitation. Stay informed about the latest advancements and best practices in the field.
Hybrid optimization of pumped hydro system and solar- Engr. Abdul-Azeez.pdffxintegritypublishin
Advancements in technology unveil a myriad of electrical and electronic breakthroughs geared towards efficiently harnessing limited resources to meet human energy demands. The optimization of hybrid solar PV panels and pumped hydro energy supply systems plays a pivotal role in utilizing natural resources effectively. This initiative not only benefits humanity but also fosters environmental sustainability. The study investigated the design optimization of these hybrid systems, focusing on understanding solar radiation patterns, identifying geographical influences on solar radiation, formulating a mathematical model for system optimization, and determining the optimal configuration of PV panels and pumped hydro storage. Through a comparative analysis approach and eight weeks of data collection, the study addressed key research questions related to solar radiation patterns and optimal system design. The findings highlighted regions with heightened solar radiation levels, showcasing substantial potential for power generation and emphasizing the system's efficiency. Optimizing system design significantly boosted power generation, promoted renewable energy utilization, and enhanced energy storage capacity. The study underscored the benefits of optimizing hybrid solar PV panels and pumped hydro energy supply systems for sustainable energy usage. Optimizing the design of solar PV panels and pumped hydro energy supply systems as examined across diverse climatic conditions in a developing country, not only enhances power generation but also improves the integration of renewable energy sources and boosts energy storage capacities, particularly beneficial for less economically prosperous regions. Additionally, the study provides valuable insights for advancing energy research in economically viable areas. Recommendations included conducting site-specific assessments, utilizing advanced modeling tools, implementing regular maintenance protocols, and enhancing communication among system components.
Welcome to WIPAC Monthly the magazine brought to you by the LinkedIn Group Water Industry Process Automation & Control.
In this month's edition, along with this month's industry news to celebrate the 13 years since the group was created we have articles including
A case study of the used of Advanced Process Control at the Wastewater Treatment works at Lleida in Spain
A look back on an article on smart wastewater networks in order to see how the industry has measured up in the interim around the adoption of Digital Transformation in the Water Industry.
Cosmetic shop management system project report.pdfKamal Acharya
Buying new cosmetic products is difficult. It can even be scary for those who have sensitive skin and are prone to skin trouble. The information needed to alleviate this problem is on the back of each product, but it's thought to interpret those ingredient lists unless you have a background in chemistry.
Instead of buying and hoping for the best, we can use data science to help us predict which products may be good fits for us. It includes various function programs to do the above mentioned tasks.
Data file handling has been effectively used in the program.
The automated cosmetic shop management system should deal with the automation of general workflow and administration process of the shop. The main processes of the system focus on customer's request where the system is able to search the most appropriate products and deliver it to the customers. It should help the employees to quickly identify the list of cosmetic product that have reached the minimum quantity and also keep a track of expired date for each cosmetic product. It should help the employees to find the rack number in which the product is placed.It is also Faster and more efficient way.
Hierarchical Digital Twin of a Naval Power SystemKerry Sado
A hierarchical digital twin of a Naval DC power system has been developed and experimentally verified. Similar to other state-of-the-art digital twins, this technology creates a digital replica of the physical system executed in real-time or faster, which can modify hardware controls. However, its advantage stems from distributing computational efforts by utilizing a hierarchical structure composed of lower-level digital twin blocks and a higher-level system digital twin. Each digital twin block is associated with a physical subsystem of the hardware and communicates with a singular system digital twin, which creates a system-level response. By extracting information from each level of the hierarchy, power system controls of the hardware were reconfigured autonomously. This hierarchical digital twin development offers several advantages over other digital twins, particularly in the field of naval power systems. The hierarchical structure allows for greater computational efficiency and scalability while the ability to autonomously reconfigure hardware controls offers increased flexibility and responsiveness. The hierarchical decomposition and models utilized were well aligned with the physical twin, as indicated by the maximum deviations between the developed digital twin hierarchy and the hardware.
Industrial Training at Shahjalal Fertilizer Company Limited (SFCL)MdTanvirMahtab2
This presentation is about the working procedure of Shahjalal Fertilizer Company Limited (SFCL). A Govt. owned Company of Bangladesh Chemical Industries Corporation under Ministry of Industries.
Planning Of Procurement o different goods and services
Yoga Posture Classification using Computer Vision
1. International Journal of Engineering and Management Research e-ISSN: 2250-0758 | p-ISSN: 2394-6962
Volume-11, Issue-4 (August 2021)
www.ijemr.net https://doi.org/10.31033/ijemr.11.4.11
86 This Work is under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
Yoga Posture Classification using Computer Vision
Madhura Prakash1
, Aishwarya S2
, Disha Maru3
, Naman Chandra4
and Varshini V5
1
Assistant Professor, Department of Information Science and Engineering, B.N.M Institute of Technology, INDIA
2
Student, Department of Information Science and Engineering, B.N.M Institute of Technology, INDIA
3
Student, Department of Information Science and Engineering, B.N.M Institute of Technology, INDIA
4
Student, Department of Information Science and Engineering, B.N.M Institute of Technology, INDIA
5
Student, Department of Information Science and Engineering, B.N.M Institute of Technology, INDIA
1
Corresponding Author: madhuraprakash5@gmail.com
ABSTRACT
There has been over the past few years, a very
increased popularity for yoga. A lot of literatures have been
published that claim yoga to be beneficial in improving the
overall lifestyle and health especially in rehabilitation,
mental health and more. Considering the fast-paced lives
that individuals live, people usually prefer to exercise or
work-out from the comfort of their homes and with that a
need for an instructor arises. Hence why, we have
developed a self-assisted system which can be used to detect
and classify yoga asanas, which is discussed in-depth in this
paper. Especially now when the pandemic has taken over
the world, it is not feasible to attend physical classes or
have an instructor over. Using the technology of Computer
Vision, a computer-assisted system such as the one
discussed, comes in very handy. The technologies such as
ml5.js, PoseNet and Neural Networks are made use for the
human pose estimation and classification. The proposed
system uses the above-mentioned technologies to take in a
real-time video input and analyze the pose of an individual,
and classifies the poses into yoga asanas. It also displays the
name of the yoga asana that is detected along with the
confidence score.
Keywords— ml5.js, Neural Network, PoseNet
I. INTRODUCTION
Human pose recognition is considered as a
well-established computer vision method has introduced
several challenges over the years. Its key function is to
locate key-points of a person's body in many fields
including video-surveillance, assisted living, human–
computer interaction, biometrics, sports, in-home health
monitoring, etc. When we consider status of the health of
a particular individual, it can be evaluated and predicted
with the help of monitoring and recognizing their
activities.
The application of Yoga posture recognition is
more or less new. Yoga, nowadays is gaining increasing
importance in the field of medical research community,
and numerous literatures have been proposed for various
medical applications such as cardiac rehabilitation,
positive body image intervention, and mental illnesses.
Yoga exercises boost physical fitness, it also helps in
cleansing the body, mind, and soul. It contains many
asanas and each of them shows the physical postures.
For people who haven't been exercising in a while, Yoga
is very helpful. It’s helpful for people who have
conditions such as arthritis or osteoporosis.
Considering the fast-paced lives, exercising at
home is mostly preferred by people these days, as some
of the resources are not publicly available, human pose
recognition can be utilized to develop a self-assist
exercise system that allows individuals to understand,
learn and practice exercises correctly by themselves. The
performance of the users by a computer-assisted self-
training systems especially in sports to prevent injuries.
These kinds of computer-assisted systems come into use
when groupwise activities and other interactions may not
be feasible when there is a pandemic. Similar works
have been proposed which includes automated and semi-
automated systems for analyzing the sports and exercise
activities which can be accessed according to one’s
convenience.
II. METHODOLOGY
CNN is used internally by PoseNet for
recognizing human poses and giving key-points as
output. The neural network is trained on key points of
joint locations of the human skeleton or can be
trained directly on the video using CNN. The model
was used on the PoseNet key points to classify yoga
asanas and achieved an average accuracy of 98%,
hence there was no need for Convolution Neural
Network.
To detect parts such as elbow hips, wrists,
knees and ankles, a deep learning TensorFlow model
called PoseNet which allows one to estimate and
track human poses is used. It is more user-friendly
and requires minimum 4GB RAM, 2GB GPU do
detect poses in real-time with good response time The
model can estimate the position of a person’s joints
in an image. The aim is to identify the positions of
feature points in order to track human body
movement. The confidence score and an array of key
points listed by Part ID, each with score and position
for all 17 points are returned. By getting (x, y)
coordinates of each joint, one can identify pose of the
person. In such a manner, continuously data is
2. International Journal of Engineering and Management Research e-ISSN: 2250-0758 | p-ISSN: 2394-6962
Volume-11, Issue-4 (August 2021)
www.ijemr.net https://doi.org/10.31033/ijemr.11.4.11
87 This Work is under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
collected for a certain Asana and stored in the Json
file. The process is repeated for all the Asanas. To
make ML approachable for a wide audience of artists,
creative coders, and students ml5.js is used. The
access to ML algorithms and models within the
browser is provided by ml5.js library, which is made
on top of TensorFlow.js with no other external
dependencies.
III. PRIOR APPROACH
There are generally three approaches for
pose recognition: OpenPose, Posenet and PifPaf.
OpenPose is a multi-person key point detection
technique that has brought a dramatic change in the
domain of pose estimation. It uses convolutional
neural network architecture that recognises facial,
foot and hand key points of an individual from an
image. The Human body joints are recognized using
the RGB camera. Here the keypoints are nose, ears,
shoulders, wrists, ankles, hips, knees, elbows, neck
and eyes. The result obtained is represented by
processing the inputs from a camera in static images,
pre-recorded videos or real time video as keypoints.
The approach proposed in [1] uses Long short-term
memory and Convolutional neural network on data
received from OpenPose that detects and extracts the
key points which are passed to the model where
convolutional neural network detects patterns and
long short-term memory analyses the change over
time. The model recognizes the six asanas from
recorded videos and in real time for twelve
individuals with an accuracy of 98.92%.
Figure 1: Existing Mechanism
The approach [2] proposes an application where
a variety of images - cat and dog are used for
classification. On CPU System, four classifier
combinations and activation functions are compared
with four different structures of convolutional neural
network. Here Keras, Tensorflow, Deep Learning, Relu,
Sigmoid, CNN, Tanh, SoftMax, Image Classification are
used. Similarly in the approach [3] uses deep learning to
classify yoga images. CMU'S open source OpenPose is
used as preprocessing module and the model uses RGB
skeleton image dataset which runs through CNN
classifier. The drawback was that the accuracy achieved
is 78% with a smaller dataset. An approach using
3DCNN method is also proposed [4] where 3DCNN
method is used in order to recognize the yoga pose
which is then introduced to supplementary layers like
batch normalization and average pooling that improves
computational efficiency. In approach [5] using Part
Affinity Fields proposes PAF to learn on how to
associate body parts with individuals from the image as
proposed in [6]. High accuracy and real time
performance is gained from the bottom-up system
regardless the number of persons within the image. The
only drawback is that the implementation is with respect
to images only. Similarly Joint Regressors [7] are used
to address the issue of estimating two- dimensional pose
from still images. As Joint Regressors two layered
random forests are employed where the first layer of
random forest acts as independent, discriminative part
classifier and the second layer of random forest
considers the estimated class distributions of the first one
into account and thereby is able to predict joint
locations. PoseNet considers input as a processed camera
image and outputs data about the keypoints. The
keypoints detected are listed by a part ID, with a
confidence score in the range of 0.0 and 1.0. The
confidence score shows the probability that a key point
can be used to estimate a single or multiple poses, which
means the version of algorithm can detect only single
individual from an image or video. The approach [8]
proposes that PoseNet grasps single 224x224 RGB
image and regresses the camera's six degrees of freedom
pose. The system instructs a CNN to regress the six
degrees of freedom cameras pose from a RGB image
until the end. The algorithm works in indoors and in
open-air in real time considering 5ms per frame to
determine. It also aims to trace the usage of multi-view
geometry as a training data source for deep pose
regressors and survey the probabilistic extensions to the
algorithm in future. PifPaf is the latest approach for two
dimensional multi person human pose estimation. It is a
bottom-up approach and uses part intensity field to
localize the body part and part association field to
associate body parts in order to structure into human
poses. With respect to lower resolution and better
performance in jam-packed places it is considered as a
better model due to the encoding of the information in a
newer composite part association field and integration of
selected Laplace Loss. The approach [9] proposes a
bottom-up method for multi person 2D human pose
estimation that is applicable for urban mobility. The goal
of the method is to estimate human poses in packed
images.
3. International Journal of Engineering and Management Research e-ISSN: 2250-0758 | p-ISSN: 2394-6962
Volume-11, Issue-4 (August 2021)
www.ijemr.net https://doi.org/10.31033/ijemr.11.4.11
88 This Work is under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
IV. OUR APPROACH
Browser friendly applications can be used by
anyone with a browser like Google Chrome or Microsoft
Edge. The aim is to make the application accessible to
everyone, even who may not have good system
requirements. For this purpose, we have used ml5.js in
our application. It is a JavaScript library built on top of
TensorFlow.js. It is a library for handling operations that
are GPU accelerated and helps in managing memory for
machine learning algorithms. Ml5.js gives direct access
to pre-trained models for detecting human poses within
the browser. The public understanding of machine
learning and fostering deep engagement with ethical
computing is given by ml5.js. Ml5.js provides
approachable easy to use APIs as one can focus on the
logic and flow rather than in-depth functions and
implementations.
As previously mentioned, the machine learning
model that allows for Real-time Human Pose Estimation
is PoseNet. There are different versions of the PoseNet
algorithm which can detect for a single pose or multiple
poses in an image or a video. For the scope of this
paper, we have considered single pose only. The
approach is divided into three parts namely, data
collection, training the model, and classification.
4.1 Data Collection
We use p5.js, a JavaScript library to capture the
video in real-time. Each video frame is loaded into the
ml5.posenet model. Ml5’s implementation of PoseNet,
uses one of the two versions, “MobileNetV1” and
“ResNet50”. These models are types of CNNs. CNN
finds pattern in the pixels of images and through
successive layers of computation finds sets of patterns to
identify more complex patterns. The model is initially
loaded before starting the data collection. On detecting
a pose, we add the collected keypoints from the pose
detected to the input array corresponding to the label
specified. The keypoints are as shown in Table 1.
The keypoints are also displayed on the
video frame. In this case, the labels indicate different
asanas of our system. The data collected is added to
the neural network post the collection and finally
saved locally as a JSON file.
4.2 Training the Model
The ml5.neuralNetwork can do classification or
regression tasks. The generalized steps for using the
ml5.neural network is as follows:
Step 1: load data or create some data
Step 2: set the neural network options &
initialize the neural network
Step 3: add data to the neural network
Step 4: normalize the data
Step 5: train the neural network
Step 6: use the trained model to make a
classification or regression
Figure 2: Epoch value set to 200
After loading the data, i.e., the previously
saved Json file with keypoints, ml5.neuralnetwork is
set with the options that indicate inputs, outputs,
task(classification). Custom layers can also be set
along with the options. The data is then normalized
using normalizeData() which normalizes the data on
a scale from 0 to 1. and the loaded data is trained
with an epoch value as shown in the Fig 1. The model
obtained is saved using save() and consists of three
files, the model i.e., model.json, the metadata i.e.,
model_meta.json, and the weight file i.e.,
model.weights.bin.
4.3 The Classification
Using the load() function of ml5.js, the three
model files are loaded into the neutral network. The
PoseNet model is loaded initially as the program
starts and the real time video input is taken using
p5.js canvas and as the PoseNet detects a pose, the
pose key points are given as input for the neural
network to be classified into one of the asanas. The
pose is classified into one of the eight asanas as
shown in the table 2. There is one default pose i.e.,
Table 1: Key Points Identified By Posenet Model
Id Part
0 Nose
1 Left Eye
2 Right Eye
3 Left Ear
4 Right Ear
5 Left Shoulder
6 Right Shoulder
7 Left Elbow
8 Right Elbow
9 Left Wrist
10 Right Wrist
11 Left Hip
12 Right Hip
13 Left Knee
14 Right Knee
15 Left Ankle
16 Right Ankle
4. International Journal of Engineering and Management Research e-ISSN: 2250-0758 | p-ISSN: 2394-6962
Volume-11, Issue-4 (August 2021)
www.ijemr.net https://doi.org/10.31033/ijemr.11.4.11
89 This Work is under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
standing, when the user is idle, this is the result
displayed.
The confidence is obtained from the results
of classify() function. A threshold of 0.75 is set for
the confidence score and the classified asana is
displayed on the screen.
V. ANALYSIS
Ml5.neuralNetwork provides classification and
regression tasks. Once the neural network is trained, it
can perform either of the tasks. Usually, classification is
used when the prediction is done for discrete values, in
this case, the different asanas are the discrete values.
Whereas, regression is performed for continuous values.
The system was tried for both the tasks and the results
have been tabulated in Table 3. The average confidence
achieved by classification was approximately 10% more
than regression.
VI. CONCLUSION
In this paper, we discuss about the Computer
Vision based system that classifies the yoga asanas along
with its confidence score, which is the measure of how
confident the machine learning model is in classifying
and identifying the asanas. The dataset is assembled
using the webcam for 3-4 persons. The keypoints of the
individuals are detected by PoseNet. The use of Neural
Networks and ml5.js on the PoseNet data is observed to
be effective and classifies all the 7+1 yoga asanas
excellently. Performance of the Convolution Neural
Networks (PoseNet) proves that Machine Learning can
also be implemented for pose estimation or activity
recognition problems. The proposed model currently
classifies only 7+1 yoga asanas. There are a number of
asanas, and hence creating a pose estimation model that
can be successful for all the asanas is a challenging
problem. The dataset can be extended by increasing the
number of yoga asanas performed by individuals and
multiple individuals as well, which works not only in an
indoor setting but also outdoors.
REFERENCES
[1] Yadav, Santosh Singh, Amitojdeep Gupta, &
Abhishek Raheja, Jagdish. (2019). Real- time Yoga
recognition using deep learning. Available at:
https://link.springer.com/article/10.1007/s00521-019-
04232-7.
[2] LI, Ang LI, Yi-xiang LI, & Xue-hui. (2017).
TensorFlow and keras-based convolutional neural
network in CAT image recognition. DEStech
Transactions on Computer Science and Engineering.
DOI: 10.12783/dtcse/cmsam2017/16428.
[3] A. Lai, B. Reddy, & B. Vlijmen. (2019). Yog.ai:
deep learning for yoga. Available at:
http://cs230.stanford.edu/projects_winter_2019/reports/1
5813480.pdf.
[4] Jain, Shrajal, Rustagi, Aditya, Saurav, Sumeet, Saini,
Ravi, & Singh, Sanjay. (2020). Three- dimensional
CNN-inspired deep learning architecture for Yoga pose
recognition in the real-world environment. Neural
Computing and Applications. DOI: 10.1007/s00521-020-
05405-5.
[5] Z. Cao, G. Hidalgo Martinez, T. Simon, S. -E. Wei, &
Y. A. Sheikh. (2019). OpenPose: Realtime multi-person
2D pose estimation using part affinity fields. In: IEEE
Transactions on Pattern Analysis and Machine
Intelligence. DOI: 10.1109/TPAMI.2019.2929257.
[6] Z. Cao, T. Simon, S. Wei, & Y. Sheikh. (2017).
Realtime multi-person 2D pose estimation using part
affinity fields. IEEE Conference on Computer Vision
and Pattern Recognition (CVPR), Honolulu, HI, pp.
1302-1310. DOI: 10.1109/CVPR.2017.143.
[7] M. Dantone, J. Gall, C. Leistner and L., & Van Gool.
(2013). Human pose estimation using body parts
dependent joint regressors. IEEE Conference on
Computer Vision and Pattern Recognition, Portland, pp.
3041-3048. DOI: 10.1109/CVPR.2013.391.
[8] A. Kendall, M. Grimes, & R. Cipolla. (2015).
PoseNet: A convolutional network for real- time 6-DOF
camera relocalization. IEEE International Conference on
Computer Vision (ICCV), Santiago, pp. 2938-2946.
DOI: 10.1109/ICCV.2015.336.
[9] S. Kreiss, L. Bertoni, & A. Alahi. (2019). PifPaf:
Composite fields for human pose estimation. IEEE/CVF
Conference on Computer Vision and Pattern
Recognition (CVPR), Long Beach, CA, USA, pp. 11969-
11978. DOI: 10.1109/CVPR.2019.01225.
Table 2: The List of Asanas
No. Asana
1 Siddhasana
2 Trikonasana
3 Vajrasana
4 Veerabhadrasana
5 Vrukshasana
6 Anjaneyasana
7 Dandasana
8 Default – Standing
Table 3: The List Of Asanas with Confidence Average
No. Asana Classifier Regressor
1 Siddhasana 99.66743 90.00125
2 Trikonasana 97.87977 87.16578
3 Vajrasana 99.38124 89.95621
4 Veerabhadrasana 99.25908 88.83210
5 Vrukshasana 98.42687 90.63447
6 Anjaneyasana 97.10453 88.45204
7 Dandasana 98.44310 87.86524
8 Default 99.56122 89.96125