InternVideo: General Video Foundation Models via Generative and Discriminative Learning
The foundation models have recently shown excellent performance on a variety of downstream tasks in computer vision. However, most existing vision foundation models simply focus on image-level pretraining and adpation, which are limited for dynamic and complex video-level understanding tasks. To fill the gap, we present general video foundation models, InternVideo, by taking advantage of both generative and discriminative self-supervised video learning. Specifically, InternVideo efficiently explores masked video modeling and video-language contrastive learning as the pretraining objectives, and selectively coordinates video representations of these two complementary frameworks in a learnable manner to boost various video applications. Without bells and whistles, InternVideo achieves state-of-the-art performance on 39 video datasets from extensive tasks including video action recognition/detection, video-language alignment, and open-world video applications. Especially, our methods can obtain 91.1% and 77.2% top-1 accuracy on the challenging Kinetics-400 and Something-Something V2 benchmarks, respectively. All of these results effectively show the generality of our InternVideo for video understanding. The code will be released at https://github.com/OpenGVLab/InternVideo .
PDF AbstractCode
Tasks
Action Classification
Action Recognition
Contrastive Learning
Open Set Action Recognition
Spatio-Temporal Action Localization
Temporal Action Localization
Video Question Answering
Video Retrieval
Video Understanding
Visual Question Answering (VQA)
Zero-Shot Video Question Answer
Zero-Shot Video Retrieval
UCF101
Kinetics
HMDB51
ActivityNet
Kinetics 400
MSR-VTT
THUMOS14
MSVD
Something-Something V2
HowTo100M
DiDeMo
WebVid
Kinetics-600
TVQA
Something-Something V1
LSMDC
VATEX
AVA
Kinetics-700
TGIF-QA
NExT-QA
HACS
TGIF
VLN-CE
FineAction
STAR Benchmark
Results from the Paper
Ranked #1 on
Action Recognition
on Something-Something V1
(using extra training data)
