Advancing High-Resolution Video-Language Representation with Large-Scale Video Transcriptions
About
We study joint video and language (VL) pre-training to enable cross-modality learning and benefit plentiful downstream VL tasks. Existing works either extract low-quality video features or learn limited text embedding, while neglecting that high-resolution videos and diversified semantics can significantly improve cross-modality learning. In this paper, we propose a novel High-resolution and Diversified VIdeo-LAnguage pre-training model (HD-VILA) for many visual tasks. In particular, we collect a large dataset with two distinct properties: 1) the first high-resolution dataset including 371.5k hours of 720p videos, and 2) the most diversified dataset covering 15 popular YouTube categories. To enable VL pre-training, we jointly optimize the HD-VILA model by a hybrid Transformer that learns rich spatiotemporal features, and a multimodal Transformer that enforces interactions of the learned video features with diversified texts. Our pre-training model achieves new state-of-the-art results in 10 VL understanding tasks and 2 more novel text-to-visual generation tasks. For example, we outperform SOTA models with relative increases of 40.4% R@1 in zero-shot MSR-VTT text-to-video retrieval task and 55.4% in high-resolution dataset LSMDC. The learned VL embedding is also effective in generating visually pleasing and semantically relevant results in text-to-visual editing and super-resolution tasks.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Video Question Answering | MSRVTT-QA | Accuracy40 | 481 | |
| Text-to-Video Retrieval | DiDeMo (test) | R@128.8 | 376 | |
| Video Question Answering | MSRVTT-QA (test) | Accuracy40 | 371 | |
| Text-to-Video Retrieval | DiDeMo | R@10.288 | 360 | |
| Text-to-Video Retrieval | MSR-VTT (test) | R@135.6 | 234 | |
| Text-to-Video Retrieval | LSMDC (test) | R@117.4 | 225 | |
| Text-to-Video Retrieval | MSR-VTT (1k-A) | R@1078 | 211 | |
| Text-to-Video Retrieval | MSRVTT (test) | Recall@10.356 | 155 | |
| Video Question Answering | TGIF-QA (test) | -- | 89 | |
| Text-to-Video Retrieval | ActivityNet-captions (val1) | R@128.5 | 58 |