2. "
Determine best language
Speed
Robustness
Portability
Available Libraries
Ease of Implementation
Experiment with different algorithms
Proto-Prototypes
4. "
Java
Library for extracting features from audio files
28 implemented features
GUI for testing
XML driven
Wide Array of formats supported
jAudio
6. "
Project Requirements pHash
Audio Fingerprinting Video,Image & Audio Fingerprinting
Substring Comparison Only a few seconds are needed for match
Speed C/C++
Runs on Lab Computers( Ubuntu) Open Source/Linux friendly
Confidence Level Bit Error Rate between hashes( 0-1)
Comparison
7. "
Compiling code on Debian/Ubuntu
Portability (sudo, etc)
Library dependencies
Developed mainly to prove thesis
Lack of documentation
Lack of widespread use (apt-cache rdepends
libphash0)
Lack of C++ knowledge
There is no Silver Bullet
15. "
Bark Scale - Looks only at frequencies that
humans hear
Frame size - 0.21 seconds
The value is based on the bark scale values of
the frame FFT spectrum
Hash computed from temporal and spectral
differences on the bark scale.
Compares Bit Error Rate between different
hashes, normalized for time.
Yeah but…
17. "
Use threads to generate hashes concurrently
Make format detection more robust
Better way of determining % of confidence
Refactor
Improvements