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

arXiv: https://arxiv.org/abs/2605.08810 · HTML: https://mlsanity.com/paper/2605.08810/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 32 of its 43 references that are in the map.
- First posted on arXiv 9 May 2026.

## Builds on (most cited first, 30 of 32)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Visual Instruction Tuning](https://mlsanity.com/paper/2304.08485/) (2023)
- [BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models](https://mlsanity.com/paper/2301.12597/) (2023)
- [Rethinking the Inception Architecture for Computer Vision](https://mlsanity.com/paper/1512.00567/) (2015)
- [Flamingo: a Visual Language Model for Few-Shot Learning](https://mlsanity.com/paper/2204.14198/) (2022)
- [Gaussian Error Linear Units (GELUs)](https://mlsanity.com/paper/1606.08415/) (2016)
- [InstructBLIP: Towards General-purpose Vision-Language Models with Instruction Tuning](https://mlsanity.com/paper/2305.06500/) (2023)
- [SlowFast Networks for Video Recognition](https://mlsanity.com/paper/1812.03982/) (2018)
- [Is Space-Time Attention All You Need for Video Understanding?](https://mlsanity.com/paper/2102.05095/) (2021)
- [VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training](https://mlsanity.com/paper/2203.12602/) (2022)
- [Self-Attentive Sequential Recommendation](https://mlsanity.com/paper/1808.09781/) (2018)
- [Video Swin Transformer](https://mlsanity.com/paper/2106.13230/) (2021)
- [Session-based Recommendations with Recurrent Neural Networks](https://mlsanity.com/paper/1511.06939/) (2015)
- [Perceiver IO: A General Architecture for Structured Inputs & Outputs](https://mlsanity.com/paper/2107.14795/) (2021)
- [Otter: A Multi-Modal Model with In-Context Instruction Tuning](https://mlsanity.com/paper/2305.03726/) (2023)
- [Ups and Downs: Modeling the Visual Evolution of Fashion Trends with One-Class Collaborative Filtering](https://mlsanity.com/paper/1602.01585/) (2016)
- [Where to Go Next for Recommender Systems? ID- vs. Modality-based Recommender Models Revisited](https://mlsanity.com/paper/2303.13835/) (2023)
- [KuaiRec: A Fully-observed Dataset and Insights for Evaluating Recommender Systems](https://mlsanity.com/paper/2202.10842/) (2022)
- [KuaiRand: An Unbiased Sequential Recommendation Dataset with Randomly Exposed Videos](https://mlsanity.com/paper/2208.08696/) (2022)
- [A Content-Driven Micro-Video Recommendation Dataset at Scale](https://mlsanity.com/paper/2309.15379/) (2023)
- [Tenrec: A Large-scale Multipurpose Benchmark Dataset for Recommender Systems](https://mlsanity.com/paper/2210.10629/) (2022)
- [Real-time Short Video Recommendation on Mobile Devices](https://mlsanity.com/paper/2208.09577/) (2022)
- [Concept-Aware Denoising Graph Neural Network for Micro-Video Recommendation](https://mlsanity.com/paper/2109.13527/) (2021)
- [KuaiSAR: A Unified Search And Recommendation Dataset](https://mlsanity.com/paper/2306.07705/) (2023)
- [DVR: Micro-Video Recommendation Optimizing Watch-Time-Gain under Duration Bias](https://mlsanity.com/paper/2208.05190/) (2022)
- [Revisiting 3D ResNets for Video Recognition](https://mlsanity.com/paper/2109.01696/) (2021)
- [A Large-scale Dataset with Behavior, Attributes, and Content of Mobile Short-video Platform](https://mlsanity.com/paper/2502.05922/) (2025)
- [Exploring the Design Space of Visual Context Representation in Video MLLMs](https://mlsanity.com/paper/2410.13694/) (2024)
- [Improving Micro-video Recommendation by Controlling Position Bias](https://mlsanity.com/paper/2208.05315/) (2022)
- [Short Video Segment-level User Dynamic Interests Modeling in Personalized Recommendation](https://mlsanity.com/paper/2504.04237/) (2025)

## Abstract (arXiv)

We propose Compressed Video Aggregator (CVA), a lightweight micro-video recommendation module that decouples video information from preference learning. CVA first summarizes frozen VFM frame embeddings into a semantic-consensus anchor through masked mean pooling, projects this anchor into a compact latent space, and refines the projected representation with residual self-attention and feedforward blocks before producing a single video embedding for the recommender. 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.