Our two papers are accepted by IEEE VCIP that will be held in Sydney Austrilia, this year! Congratulations to Huang Chao, Tong Chen, Haojie Liu, and Linfeng Liu.
One paper discusses the possbility to do extreme image compression via Generative Adversarial Optimization. This gives us another view about the loss function design in an end-to-end learning based image compression framework. The other one studies the retrieval tasks using compressed bitstream, where compression method is based on our proposed neural image coding. This is a follow-up work after our CodedVision where we explored the possibility to do classification in a compressive way. Both papers are trying to explore new avenues in the line of nerual image/video communications.