Compressed Video Aggregator: Content-driven Module for Efficient Micro-Video Recommendation

TL;DR

The paper introduces CVA, a lightweight micro-video recommendation module that efficiently aggregates video embeddings, reducing training time and memory while maintaining or improving performance.

Contribution

It presents a novel content-driven approach that decouples video information from preference learning, with frame re-selection based on CLIP to enhance recommendation accuracy.

Findings

CVA achieves consistent performance gains on MicroLens and Short-Video datasets.

Re-selected key frames further improve recommendation performance.

Significant reductions in training time and GPU memory usage.

Abstract

We propose Compressed Video Aggregator (CVA), a lightweight micro-video recommendation module that decouples video information from preference learning. It aggregates frozen VFM embeddings, and uses latent reasoning without cross-attention projection, producing compact video embeddings for recommenders. Due to the redundancy in the frame count of the original benchmark and its overly coarse sampling, we used titles to re-select key frames based on CLIP. Experiments on MicroLens and Short-Video show consistent gains with orders-of-magnitude reductions in training time and GPU memory, and re-selected frames can further enhance the performance of all methods, including CVA. Furthermore, we also discussed the impact of several scenarios involving erroneous titles on our method. Code will be released soon.