Research map / Video / Action recognition
Mode Variational LSTM Robust to Unseen Modes of Variation: Application to Facial Expression Recognition
Wissam J. Baddar, Yong Man Ro
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 |
|---|---|---|---|
| Add a separate bias cell and static-sequence path to the LSTM | LSTM · Oulu-CASIA | recognition rate (%): 78.21 -> 83.94 | Improved |
| Add a separate bias cell and static-sequence path to the LSTM | LSTM · KAIST Face MPMI | recognition rate (%): 79.93 -> 82.61 | Improved |
| Add a separate bias cell and static-sequence path to the LSTM | LSTM · AFEW validation | visual recognition rate (%): 48.83 vs 48.60 | Improved |
| Add cross-cell peephole connections between the dynamics and bias cells | Mode variational LSTM · Oulu-CASIA | recognition rate (%): 83.94 -> 85.18 | Improved |
| Add cross-cell peephole connections between the dynamics and bias cells | Mode variational LSTM · AFEW validation | visual recognition rate (%): 48.83 -> 51.44 | Improved |
| Add cross-cell peephole connections between the dynamics and bias cells | Mode variational LSTM · KAIST Face MPMI | recognition rate (%): 82.61 -> 84.98 | Improved |
| Add a separate bias cell and static-sequence path | LSTM · Oulu-CASIA cross-race, Asian train to Caucasian test | recognition rate (%): 73.39 -> 79.14 | Improved |
| Add a separate bias cell and static-sequence path | LSTM · Oulu-CASIA cross-race, Caucasian train to Asian test | recognition rate (%): 74.67 -> 80.95 | Improved |
| Use the mode variational LSTM with cross-cell peephole | LSTM · KAIST Face MPMI, room illumination | recognition rate (%): 78.21 -> 83.11 | Improved |
| Use the mode variational LSTM with cross-cell peephole | LSTM · KAIST Face MPMI, bright illumination | recognition rate (%): 76.98 -> 83.03 | Improved |
| Use the mode variational LSTM with cross-cell peephole | LSTM · KAIST Face MPMI, left illumination | recognition rate (%): 73.67 -> 82.87 | Improved |
| Use the mode variational LSTM with cross-cell peephole | LSTM · KAIST Face MPMI, right illumination | recognition rate (%): 74.39 -> 82.91 | Improved |
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.
- Lomo — LOMo: Latent Ordinal Model for Facial Analysis in Videos (CVPR 2016) · better on 1 of 1 shared result
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
- EmWorld: Emotion World Model with Latent State Evolution for Scenario-Incremental Dynamic Facial Expression Recognition (ICML 2026 · similarity 0.85)
- A Novel Dynamic Model Capturing Spatial and Temporal Patterns for Facial Expression Analysis (TPAMI 2019 · similarity 0.85)
- Learning Spatio-Temporal Features With Partial Expression Sequences for On-the-Fly Prediction (AAAI 2018 · similarity 0.85)
- Lifting Scheme-Based Implicit Disentanglement of Emotion-Related Facial Dynamics in the Wild (AAAI 2025 · similarity 0.83)
- Learning Adaptive Spatial Coherent Correlations for Speech-Preserving Facial Expression Manipulation (CVPR 2024 · similarity 0.83)
- Navigating Label Ambiguity for Facial Expression Recognition in the Wild (AAAI 2025 · similarity 0.83)
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.
