InvKA: Gait Recognition via Invertible Koopman Autoencoder
This work addresses gait recognition for applications like surveillance or biometrics, offering a domain-specific improvement with competitive performance and significant efficiency gains.
The paper tackled the problems of poor interpretability and high computational cost in gait recognition by introducing an invertible Koopman autoencoder that uses Koopman operator theory to capture kinematic features, achieving 98% accuracy on non-occlusion datasets and reducing computational cost to 1% compared to state-of-the-art methods.
Most current gait recognition methods suffer from poor interpretability and high computational cost. To improve interpretability, we investigate gait features in the embedding space based on Koopman operator theory. The transition matrix in this space captures complex kinematic features of gait cycles, namely the Koopman operator. The diagonal elements of the operator matrix can represent the overall motion trend, providing a physically meaningful descriptor. To reduce the computational cost of our algorithm, we use a reversible autoencoder to reduce the model size and eliminate convolutional layers to compress its depth, resulting in fewer floating-point operations. Experimental results on multiple datasets show that our method reduces computational cost to 1% compared to state-of-the-art methods while achieving competitive recognition accuracy 98% on non-occlusion datasets.