Summary of Learning to Compress Contexts For Efficient Knowledge-based Visual Question Answering, by Weixi Weng et al.
Learning to Compress Contexts for Efficient Knowledge-based Visual Question Answering
by Weixi Weng, Jieming Zhu, Xiaojun Meng, Hao Zhang, Rui Zhang, Chun Yuan
First submitted to arxiv on: 11 Sep 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Machine Learning (cs.LG)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The paper proposes Retrieval-Augmented Multimodal Large Language Models with Compressed Contexts (RACC) to address the issue of decreasing inference efficiency when dealing with increasing input tokens in Knowledge-Based Visual Question Answering (KB-VQA). RACC learns to compress and aggregate retrieved knowledge for a given image-question pair, generating a compact modulation in the form of Key-Value cache to adapt the downstream frozen MLLM. This approach achieves state-of-the-art performance on OK-VQA with an accuracy of 63.92% and significantly reduces inference latency by 22.0%-59.7% compared to RAVQA-v2. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper is about a new way to make computer models better at answering questions that combine pictures and words. The problem is that when you give the model more information, it gets slower. So, the researchers came up with an idea called Retrieval-Augmented Multimodal Large Language Models with Compressed Contexts (RACC). This helps the model work faster while still being very good at answering questions. |
Keywords
» Artificial intelligence » Inference » Question answering
