Bikes
Abstract
Light field (LF) imaging captures both spatial and angular information, which is essential for applications such as depth estimation, view synthesis, and post-capture refocusing. However, the efficient processing of this data, particularly in terms of compression and execution speed, presents challenges. We propose a Variational Autoencoder (VAE)-based framework to disentangle the spatial and angular features of light field images, focusing on fast and efficient compression. Our method uses two separate sub-encoders—one for spatial and one for angular features—to allow for independent processing in the latent space, which facilitates more streamlined processing. Evaluations on standard light field datasets show that our approach reduces execution time significantly, with a slight trade-off in Rate-Distortion (RD) performance, making it suitable for real-time applications. This framework also supports enhancements in other light field processing tasks, such as view synthesis, scene reconstruction, and super-resolution.
Rate-Distortion (RD) Graph
Rate-Distortion (RD) curve comparing the performance of our proposed VAE-based light field compression method against HEVC, RLVC, HLVC, JPEG-Pleno, and Shi’s method . Our method consistently outperforms the others across most datasets in terms of compression efficiency and image reconstruction quality, except for Shi’s method, which shows superior results on a few specific datasets.
Encoding and Decoding Time
Comparison of the Encoding and Decoding time for our method VS RLVC, HLVC, Jpeg-Pleno and Shi's
