Research map / Video / Action recognition
Moments in Time Dataset: One Million Videos for Event Understanding
Mathew Monfort, Carl Vondrick, Aude Oliva, Alex Andonian, Bolei Zhou, Kandan Ramakrishnan, Sarah Adel Bargal, Tom Yan, Lisa Brown, Quanfu Fan, Dan Gutfreund
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| initialize ResNet50 from ImageNet instead of scratch | ResNet50 spatial · Moments in Time (val) | top-1 accuracy: 23.65 -> 27.16 | Improved |
| initialize ResNet50 from Places instead of scratch | ResNet50 spatial · Moments in Time (val) | top-1 accuracy: 23.65 -> 26.44 | Improved |
| ImageNet init instead of Places init | ResNet50 spatial · Moments in Time (val) | top-1 accuracy: 26.44 -> 27.16 | Improved |
| add optical flow stream to TSN spatial stream (two-stream) | TSN · Moments in Time (val) | top-1 accuracy: 24.11 -> 25.32 | Improved |
| flow modality alone vs spatial for TSN | TSN-Flow · Moments in Time (val) | top-1 accuracy: 24.11 -> 15.71 | Hurt |
| use TSN over plain BNInception on flow input | BNInception-Flow · Moments in Time (val) | top-1 accuracy: 11.60 -> 15.71 | Improved |
| inflate 3D ResNet50 from best 2D ResNet50 (I3D) rather than 2D spatial model | ResNet50 · Moments in Time (val) | top-1 accuracy: 27.16 -> 29.51 | Improved |
| ensemble spatial, spatiotemporal and auditory features with linear SVM | I3D · Moments in Time (val) | top-1 accuracy: 29.51 -> 31.16 | Improved |
| pretrain on Moments in Time instead of Kinetics | ResNet50 I3D · HMDB51 | top-1 accuracy: 62.0 -> 65.9 | Improved |
| pretrain on Moments in Time instead of Kinetics | ResNet50 I3D · Something-Something | top-1 accuracy: 48.6 -> 50.0 | Improved |
| pretrain on Moments in Time instead of Kinetics | ResNet50 I3D · UCF101 | top-1 accuracy: 92.6 -> 91.9 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- SoundNet — SoundNet: Learning Sound Representations from Unlabeled Video (NeurIPS 2016) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- About Time: Advances, Challenges, and Outlooks of Action Understanding (IJCV 2025 · similarity 0.90)
- AVA: A Video Dataset of Spatio-Temporally Localized Atomic Visual Actions (CVPR 2018 · similarity 0.90)
- MomentSeeker: A Task-Oriented Benchmark For Long-Video Moment Retrieval (NeurIPS 2025 · similarity 0.88)
- TIM: A Time Interval Machine for Audio-Visual Action Recognition (CVPR 2024 · similarity 0.88)
- Adapting to the Unknown: Training-Free Audio-Visual Event Perception with Dynamic Thresholds (CVPR 2025 · similarity 0.87)
- VidEvent: A Large Dataset for Understanding Dynamic Evolution of Events in Videos (AAAI 2025 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
