RL-RC-DoT: A Block-level RL agent for Task-Aware Video Compression
This work addresses the need for efficient task-aware video compression in applications like autonomous driving, where videos are processed by AI systems, though it is incremental as it builds on existing encoder frameworks.
The paper tackles the problem of optimizing video compression for downstream AI tasks rather than human perception by controlling quantization parameters at the macro-block level using reinforcement learning. It demonstrates significant improvements in tasks like car detection and ROI encoding, achieving better task performance for a given bit rate compared to traditional methods.
Video encoders optimize compression for human perception by minimizing reconstruction error under bit-rate constraints. In many modern applications such as autonomous driving, an overwhelming majority of videos serve as input for AI systems performing tasks like object recognition or segmentation, rather than being watched by humans. It is therefore useful to optimize the encoder for a downstream task instead of for perceptual image quality. However, a major challenge is how to combine such downstream optimization with existing standard video encoders, which are highly efficient and popular. Here, we address this challenge by controlling the Quantization Parameters (QPs) at the macro-block level to optimize the downstream task. This granular control allows us to prioritize encoding for task-relevant regions within each frame. We formulate this optimization problem as a Reinforcement Learning (RL) task, where the agent learns to balance long-term implications of choosing QPs on both task performance and bit-rate constraints. Notably, our policy does not require the downstream task as an input during inference, making it suitable for streaming applications and edge devices such as vehicles. We demonstrate significant improvements in two tasks, car detection, and ROI (saliency) encoding. Our approach improves task performance for a given bit rate compared to traditional task agnostic encoding methods, paving the way for more efficient task-aware video compression.