Semantic-assisted image compression
This addresses the issue for AI applications requiring compressed images, though it appears incremental as it builds on existing compression methods with semantic enhancements.
The paper tackles the problem of conventional image compression methods ignoring downstream AI task performance by proposing a Semantic-Assisted Image Compression (SAIC) method that maintains semantic-level consistency, achieving better performance on downstream tasks compared to traditional and advanced methods at the same compression ratio.
Conventional image compression methods typically aim at pixel-level consistency while ignoring the performance of downstream AI tasks.To solve this problem, this paper proposes a Semantic-Assisted Image Compression method (SAIC), which can maintain semantic-level consistency to enable high performance of downstream AI tasks.To this end, we train the compression network using semantic-level loss function. In particular, semantic-level loss is measured using gradient-based semantic weights mechanism (GSW). GSW directly consider downstream AI tasks' perceptual results. Then, this paper proposes a semantic-level distortion evaluation metric to quantify the amount of semantic information retained during the compression process. Experimental results show that the proposed SAIC method can retain more semantic-level information and achieve better performance of downstream AI tasks compared to the traditional deep learning-based method and the advanced perceptual method at the same compression ratio.