Appearance-based Gesture recognition in the compressed domain
This work addresses gesture recognition for applications like human-computer interaction, but it appears incremental as it builds on existing compressed domain and DTW-based methods.
The paper tackles gesture recognition by extracting features directly from compressed image measurements, improving computational efficiency and memory usage of previous DTW-based K-NN classifiers, with results supported by simulation and hardware implementation.
We propose a novel appearance-based gesture recognition algorithm using compressed domain signal processing techniques. Gesture features are extracted directly from the compressed measurements, which are the block averages and the coded linear combinations of the image sensor's pixel values. We also improve both the computational efficiency and the memory requirement of the previous DTW-based K-NN gesture classifiers. Both simulation testing and hardware implementation strongly support the proposed algorithm.