BlockCNN: A Deep Network for Artifact Removal and Image Compression
This work addresses image quality and storage issues for users of JPEG compression, but it appears incremental as it builds on existing JPEG methods with a novel network application.
The paper tackles the dual problems of JPEG artifact removal and image compression by using a deep network that processes 8x8 blocks sequentially, predicting intensities and storing residuals, while reusing JPEG routines. It reports significant performance improvements in both tasks, though no concrete numbers are provided.
We present a general technique that performs both artifact removal and image compression. For artifact removal, we input a JPEG image and try to remove its compression artifacts. For compression, we input an image and process its 8 by 8 blocks in a sequence. For each block, we first try to predict its intensities based on previous blocks; then, we store a residual with respect to the input image. Our technique reuses JPEG's legacy compression and decompression routines. Both our artifact removal and our image compression techniques use the same deep network, but with different training weights. Our technique is simple and fast and it significantly improves the performance of artifact removal and image compression.