MINI PROJECT WITH SEMINAR PPT S.Kausalya
Reg No:1951002
It is a state-of-the-art,
real-time object detection
system
TYPES OF DL ALGORITHMS
Deep Neural networks
CNN
R-CNN
Fast R-CNN
YOLO
HIDDEN LAYER IN NN
YOLO
How it works?
 Prior detection systems repurpose classifiers or
localizers to perform detection. They apply the model
to an image at multiple locations and scales. High
scoring regions of the image are considered
detections.
 We use a totally different approach. We apply a
single neural network to the full image. This network
divides the image into regions and predicts bounding
boxes and probabilities for each region. These
bounding boxes are weighted by the predicted
probabilities.
High level idea
Compared to other region proposal classification networks (fast RCNN) which
perform detection on various region proposals and thus end up performing
prediction multiple times for various regions in a image, Yolo architecture is more
like FCNN (fully convolutional neural network) and passes the image (nxn) once
through the FCNN and output is (mxm) prediction. This the architecture is splitting
the input image in mxm grid and for each grid generation 2 bounding boxes and
class probabilities for those bounding boxes. Note that bounding box is more
likely to be larger than the grid itself
Yolo
Yolo
Yolo
Yolo

Yolo

  • 1.
    MINI PROJECT WITHSEMINAR PPT S.Kausalya Reg No:1951002
  • 2.
    It is astate-of-the-art, real-time object detection system
  • 8.
    TYPES OF DLALGORITHMS Deep Neural networks CNN R-CNN Fast R-CNN YOLO
  • 10.
  • 13.
  • 15.
    How it works? Prior detection systems repurpose classifiers or localizers to perform detection. They apply the model to an image at multiple locations and scales. High scoring regions of the image are considered detections.  We use a totally different approach. We apply a single neural network to the full image. This network divides the image into regions and predicts bounding boxes and probabilities for each region. These bounding boxes are weighted by the predicted probabilities.
  • 16.
    High level idea Comparedto other region proposal classification networks (fast RCNN) which perform detection on various region proposals and thus end up performing prediction multiple times for various regions in a image, Yolo architecture is more like FCNN (fully convolutional neural network) and passes the image (nxn) once through the FCNN and output is (mxm) prediction. This the architecture is splitting the input image in mxm grid and for each grid generation 2 bounding boxes and class probabilities for those bounding boxes. Note that bounding box is more likely to be larger than the grid itself