Research map / Video / Action recognition
Discriminatively Trained Latent Ordinal Model for Video Classification
Karan Sikka, Gaurav Sharma
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 |
|---|---|---|---|
| replace single-segment MIL with LOMo ordered sub-events | MIL · Oulu-CASIA VIS (SIFT) | accuracy: 69.8 -> 74.0 | Improved |
| replace MIL with LOMo | MIL · CK+ and UNBC McMaster | accuracy: 90.8 -> 92.0; 85.9 -> 87.0 | Improved |
| replace MIL with LOMo | MIL · LILiR questioning | accuracy: 80.7 -> 86.6 | Improved |
| replace max pooling with LOMo | Max pooling (MxP) · LILiR agreeing | accuracy: 85.5 -> 79.4 | Hurt |
| add the ordinal component to LOMo (ablation) | LOMo without ordinal (ord=0) · Olympic (iDT-64-FV) | mAP: 75.8 -> 84.6 | Improved |
| add the ordinal component to LOMo (ablation) | LOMo without ordinal (ord=0) · Olympic (CNN) | mAP: 70.0 -> 69.8 | No clear effect |
| add the ordinal component to LOMo (ablation) | LOMo without ordinal (ord=0) · HighFive (iDT-64-FV) | mAP: 54.4 -> 58.1 | Improved |
| add the ordinal component to LOMo (ablation) | LOMo without ordinal (ord=0) · HMDB (iDT-64-FV) | mAP: 45.1 -> 48.2 | Improved |
| add global temporal pooling information adaptively to LOMo | LOMo · Olympic (iDT-64-FV) | mAP: 84.6 -> 88.3 | Improved |
| add global temporal pooling information adaptively to LOMo | LOMo · HMDB (iDT-64-FV) | mAP: 48.2 -> 52.2 | Improved |
| use Adaptive LOMo instead of simple MIL+GTP combination | MIL+GTP · HMDB (iDT-64-FV) | mAP: 54.1 -> 52.2 | Hurt |
| use motion-aware iDT features instead of CNN features | Adaptive LOMo vs GTP · Olympic | relative improvement over GTP: 2.1% -> 3.4% | Improved |
Closest papers by meaning
- LOMo: Latent Ordinal Model for Facial Analysis in Videos (CVPR 2016 · similarity 0.94)
- Learning Facial Action Units From Web Images With Scalable Weakly Supervised Clustering (CVPR 2018 · similarity 0.86)
- Uncertainty-Aware Weakly Supervised Action Detection from Untrimmed Videos (ECCV 2020 · similarity 0.86)
- Large-Scale Weakly-Supervised Pre-Training for Video Action Recognition (CVPR 2019 · similarity 0.86)
- Weakly Supervised Action Learning With RNN Based Fine-To-Coarse Modeling (CVPR 2017 · similarity 0.86)
- From Pixels to Semantics: Unified Facial Action Representation Learning for Micro-Expression Analysis (ICLR 2026 · similarity 0.85)
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.
