The Surprising Effectiveness of Multimodal Large Language Models for Video Moment Retrieval

• Authors: Boris Meinardus, Anil Batra, Anna Rohrbach, Marcus Rohrbach
• Paper: arxiv

We introduce Mr. BLIP (Mr. as in Moment Retrieval), a multimodal, single-stage model that requires no expensive video-language pretraining, no additional input signal (e.g., no transcript or audio) and has a simpler and more versatile design than prior state-of-the-art methods. We achieve a new state-of-the-art in moment retrieval on the widely used benchmarks Charades-STA, QVHighlights, and ActivityNet Captions. Notably, we attain over 9% (absolute) higher Recall (at 0.5 and 0.7 IoU) on the challenging long-video multi-moment QVHighlights benchmark.

Code structure

# data & data preprocessing
./mr_BLIP_data

# pretrained checkpoints
./mr_BLIP_checkpoints

# mr_BLIP code
./lavis/

# running scripts for mr_BLIP training and inference
./run_scripts

Setup

Install Dependencies

1. (Optional) Creating conda environment

conda create -n mrBlip python=3.8
conda activate mrBlip

2. build from source

pip install -r requirements.txt

Download Pretrained Models

We train Mr. BLIP on QVHighlights, Charades-STA, and ActivityNet Captions and provide the checkpoints. Download the checkpoints and put them under /mr_BLIP_checkpoints.

Dataset Preparation

We test our model on:

• Charades-STA

  • Charades (Videos)

• QVHighlights

• ActivityNet Captions

Please download original MR data and preprocess them via our scripts.

Training and Inference

We provide Mr. BLIP training and inference script examples as follows.

And please refer to dataset page to customize your data path.

You might want to update the config files for the respective runs to fit on your machine. They are currently set to run on 8 A100-80GB GPUs. You can simply reduce the batch size, reduce the number of frames, or apply a frame level embeddings aggregation (32 frame tokens -> 1 token) to fit on a smaller GPU.

1) QVH Finetuning

sh run_scripts/mr_BLIP/train/qvh.sh

2) Charades-STA Finetuning

sh run_scripts/mr_BLIP/train/charades.sh

3) ANet Captions Finetuning

sh run_scripts/mr_BLIP/train/anet.sh

4) QVH Evaluation

sh run_scripts/mr_BLIP/eval/qvh.sh

5) Charades-STA Evaluation

sh run_scripts/mr_BLIP/eval/charades.sh

6) ANet Captions Evaluation

sh run_scripts/mr_BLIP/eval/anet.sh

Acknowledgments

We thank the developers of LAVIS and BLIP-2 for their public code release.

Reference

Please cite our paper if you use our models in your works:

@article{boris2024surprisingeffectivenessmultimodallarge,
      title={The Surprising Effectiveness of Multimodal Large Language Models for Video Moment Retrieval}, 
      author={Meinardus Boris and Batra Anil and Rohrbach Anna and Rohrbach Marcus},
      year={2024},
      eprint={2406.18113},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2406.18113}, 
}