# Video-level A/H Recognition on BAH (dev val) Source: [Video-level A/H Recognition on BAH (dev val)]() Collection index: [llms.txt]() Data updated: 2026-03-19T07:44:50.734Z Evaluation of Video-level Action/Human (A/H) Recognition performance on the development/validation split of the BAH dataset, measured using the MF1 metric. A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | Video-level A/H Recognition | | Dataset (including any stated split) | BAH (dev val) | | Metric | MF1 | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 31 of 31 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [MF1](): Higher is better; unit: PERCENT. ## Recorded results | Method | MF1 | Result date | Report paper | | --- | --- | --- | --- | | Multimodal Fusion Model | 85.38 | Not recorded | 69b78e6d0636e5bf4afbdb04 | | Multimodal Fusion Model with Prototype Head | 83.79 | Not recorded | 69b78e6d0636e5bf4afbdb04 | | Ensemble of Five Multimodal Fusion Models with Prototype Head | 83 | Not recorded | 69b78e6d0636e5bf4afbdb04 | | Ensemble of Five Multimodal Fusion Models | 81.94 | Not recorded | 69b78e6d0636e5bf4afbdb04 | | Multimodal Fusion Model | 81.77 | Not recorded | 69b78e6d0636e5bf4afbdb04 | | Multimodal Fusion Model | 80.79 | Not recorded | 69b78e6d0636e5bf4afbdb04 | | Multimodal Fusion Model | 79.89 | Not recorded | 69b78e6d0636e5bf4afbdb04 | | Multimodal Fusion Model | 79.89 | Not recorded | 69b78e6d0636e5bf4afbdb04 | | Multimodal Fusion Model | 78.35 | Not recorded | 69b78e6d0636e5bf4afbdb04 | | Multimodal Fusion Model | 78.07 | Not recorded | 69b78e6d0636e5bf4afbdb04 | | Multimodal Fusion Model | 77.91 | Not recorded | 69b78e6d0636e5bf4afbdb04 | | Multimodal Fusion Model | 77.37 | Not recorded | 69b78e6d0636e5bf4afbdb04 | | Multimodal Fusion Model | 76.1 | Not recorded | 69b78e6d0636e5bf4afbdb04 | | Fine-tuned EmotionTextClassifier | 69.28 | Not recorded | 69b78e6d0636e5bf4afbdb04 | | Fusion B (divergence) | 69.12 | Not recorded | 69bb82e30636e5bf4afc181b | | Fine-tuned EmotionDistilRoBERTa | 68.54 | Not recorded | 69b78e6d0636e5bf4afbdb04 | | TF-IDF | 68.3 | Not recorded | 69b78e6d0636e5bf4afbdb04 | | Multimodal Fusion Model | 68.08 | Not recorded | 69b78e6d0636e5bf4afbdb04 | | Fusion A (implicit) | 67.88 | Not recorded | 69bb82e30636e5bf4afc181b | | EmotionWav2Vec2.0 + Mamba | 67.2 | Not recorded | 69b78e6d0636e5bf4afbdb04 | | Multimodal Fusion Model | 67.05 | Not recorded | 69b78e6d0636e5bf4afbdb04 | | Fusion C (combined) | 67 | Not recorded | 69bb82e30636e5bf4afc181b | | TF-IDF | 65.56 | Not recorded | 69b78e6d0636e5bf4afbdb04 | | EmotionEfficientNetB0 + Statistical Features | 65.29 | Not recorded | 69b78e6d0636e5bf4afbdb04 | | Multimodal Fusion Model | 65.29 | Not recorded | 69b78e6d0636e5bf4afbdb04 | | Fusion B (divergence) | 65.24 | Not recorded | 69bb82e30636e5bf4afc181b | | Multimodal Fusion Model | 63.36 | Not recorded | 69b78e6d0636e5bf4afbdb04 | | Visual AUs (XGBoost) | 61.94 | Not recorded | 69bb82e30636e5bf4afc181b | | VideoMAE | 61.71 | Not recorded | 69b78e6d0636e5bf4afbdb04 | | Text BERT | 57.58 | Not recorded | 69bb82e30636e5bf4afc181b | | Audio Wav2Vec (LSTM) | 52.18 | Not recorded | 69bb82e30636e5bf4afc181b | ### Multimodal Fusion Model MF1: 85.38. Result ID: 69b7d37747f34ce9249ffd62. [Reporting paper]() Recorded evaluation configuration: {"Modality":"Models IDs 1, 2, 3 and 7","Classifier":"Linear Layer"}. This record does not assert independent reproduction. ### Multimodal Fusion Model with Prototype Head MF1: 83.79. Result ID: 69b7d37747f34ce9249ffd63. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Modality":"Models IDs 1, 2, 3 and 7","Classifier":"Linear Layer"}. This record does not assert independent reproduction. ### Ensemble of Five Multimodal Fusion Models with Prototype Head MF1: 83. Result ID: 69b7d37747f34ce9249ffd65. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Modality":"Models IDs 1, 2, 3 and 7","Classifier":"Linear Layer"}. This record does not assert independent reproduction. ### Ensemble of Five Multimodal Fusion Models MF1: 81.94. Result ID: 69b7d37747f34ce9249ffd64. [Reporting paper]() Recorded evaluation configuration: {"Modality":"Models IDs 1, 2, 3 and 7","Classifier":"Linear Layer"}. This record does not assert independent reproduction. ### Multimodal Fusion Model MF1: 81.77. Result ID: 69b7d37747f34ce9249ffd6b. [Reporting paper]() Recorded evaluation configuration: {"Modality":"Models IDs 2 and 7","Classifier":"Linear Layer"}. This record does not assert independent reproduction. ### Multimodal Fusion Model MF1: 80.79. Result ID: 69b7d37747f34ce9249ffd5f. [Reporting paper]() Recorded evaluation configuration: {"Modality":"Models IDs 1, 2, 3 and 4","Classifier":"Linear Layer"}. This record does not assert independent reproduction. ### Multimodal Fusion Model MF1: 79.89. Result ID: 69b7d37747f34ce9249ffd6c. [Reporting paper]() Recorded evaluation configuration: {"Modality":"Models IDs 2, 3 and 7","Classifier":"Linear Layer"}. This record does not assert independent reproduction. ### Multimodal Fusion Model MF1: 79.89. Result ID: 69b7d37747f34ce9249ffd6d. [Reporting paper]() Recorded evaluation configuration: {"Modality":"Models IDs 1, 2 and 7","Classifier":"Linear Layer"}. This record does not assert independent reproduction. ### Multimodal Fusion Model MF1: 78.35. Result ID: 69b7d37747f34ce9249ffd61. [Reporting paper]() Recorded evaluation configuration: {"Modality":"Models IDs 1, 2, 3 and 6","Classifier":"Linear Layer"}. This record does not assert independent reproduction. ### Multimodal Fusion Model MF1: 78.07. Result ID: 69b7d37747f34ce9249ffd68. [Reporting paper]() Recorded evaluation configuration: {"Modality":"Models IDs 1 and 2","Classifier":"Linear Layer"}. This record does not assert independent reproduction. ### Multimodal Fusion Model MF1: 77.91. Result ID: 69b7d37747f34ce9249ffd60. [Reporting paper]() Recorded evaluation configuration: {"Modality":"Models IDs 1, 2, 3 and 5","Classifier":"Linear Layer"}. This record does not assert independent reproduction. ### Multimodal Fusion Model MF1: 77.37. Result ID: 69b7d37747f34ce9249ffd6a. [Reporting paper]() Recorded evaluation configuration: {"Modality":"Models IDs 2 and 3","Classifier":"Linear Layer"}. This record does not assert independent reproduction. ### Multimodal Fusion Model MF1: 76.1. Result ID: 69b7d37747f34ce9249ffd6e. [Reporting paper]() Recorded evaluation configuration: {"Modality":"Models IDs 1, 2 and 3","Classifier":"Linear Layer"}. This record does not assert independent reproduction. ### Fine-tuned EmotionTextClassifier MF1: 69.28. Result ID: 69b7d37747f34ce9249ffd5d. [Reporting paper]() Recorded evaluation configuration: {"Modality":"Text","Classifier":"MLP"}. This record does not assert independent reproduction. ### Fusion B (divergence) MF1: 69.12. Result ID: 69bba95547f34ce924a53c0b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Modality":"Multimodal","Input Features":"windowed AUs","Fusion Strategy":"Divergence"}. This record does not assert independent reproduction. ### Fine-tuned EmotionDistilRoBERTa MF1: 68.54. Result ID: 69b7d37747f34ce9249ffd5e. [Reporting paper]() Recorded evaluation configuration: {"Modality":"Text","Classifier":"MLP"}. This record does not assert independent reproduction. ### TF-IDF MF1: 68.3. Result ID: 69b7d37747f34ce9249ffd5b. [Reporting paper]() Recorded evaluation configuration: {"Modality":"Text","Classifier":"Logistic Regression"}. This record does not assert independent reproduction. ### Multimodal Fusion Model MF1: 68.08. Result ID: 69b7d37747f34ce9249ffd6f. [Reporting paper]() Recorded evaluation configuration: {"Modality":"Models IDs 1, 3 and 7","Classifier":"Linear Layer"}. This record does not assert independent reproduction. ### Fusion A (implicit) MF1: 67.88. Result ID: 69bba95547f34ce924a53c08. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Modality":"Multimodal","Input Features":"raw AUs","Fusion Strategy":"Implicit"}. This record does not assert independent reproduction. ### EmotionWav2Vec2.0 + Mamba MF1: 67.2. Result ID: 69b7d37747f34ce9249ffd5a. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Modality":"Audio","Classifier":"Linear Layer"}. This record does not assert independent reproduction. ### Multimodal Fusion Model MF1: 67.05. Result ID: 69b7d37747f34ce9249ffd69. [Reporting paper]() Recorded evaluation configuration: {"Modality":"Models IDs 3 and 7","Classifier":"Linear Layer"}. This record does not assert independent reproduction. ### Fusion C (combined) MF1: 67. Result ID: 69bba95547f34ce924a53c0a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Modality":"Multimodal","Input Features":"raw AUs","Fusion Strategy":"Combined"}. This record does not assert independent reproduction. ### TF-IDF MF1: 65.56. Result ID: 69b7d37747f34ce9249ffd5c. [Reporting paper]() Recorded evaluation configuration: {"Modality":"Text","Classifier":"CatBoost"}. This record does not assert independent reproduction. ### EmotionEfficientNetB0 + Statistical Features MF1: 65.29. Result ID: 69b7d37747f34ce9249ffd58. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Modality":"Face","Classifier":"MLP"}. This record does not assert independent reproduction. ### Multimodal Fusion Model MF1: 65.29. Result ID: 69b7d37747f34ce9249ffd67. [Reporting paper]() Recorded evaluation configuration: {"Modality":"Models IDs 1 and 7","Classifier":"Linear Layer"}. This record does not assert independent reproduction. ### Fusion B (divergence) MF1: 65.24. Result ID: 69bba95547f34ce924a53c09. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Modality":"Multimodal","Input Features":"raw AUs","Fusion Strategy":"Divergence"}. This record does not assert independent reproduction. ### Multimodal Fusion Model MF1: 63.36. Result ID: 69b7d37747f34ce9249ffd66. [Reporting paper]() Recorded evaluation configuration: {"Modality":"Models IDs 1 and 3","Classifier":"Linear Layer"}. This record does not assert independent reproduction. ### Visual AUs (XGBoost) MF1: 61.94. Result ID: 69bba95547f34ce924a53c05. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Modality":"Unimodal","Model":"XGBoost","Features":"Visual AUs"}. This record does not assert independent reproduction. ### VideoMAE MF1: 61.71. Result ID: 69b7d37747f34ce9249ffd59. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Modality":"Scene","Classifier":"Linear Layer"}. This record does not assert independent reproduction. ### Text BERT MF1: 57.58. Result ID: 69bba95547f34ce924a53c07. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Modality":"Unimodal","Model":"BERT-base","Features":"Text"}. This record does not assert independent reproduction. ### Audio Wav2Vec (LSTM) MF1: 52.18. Result ID: 69bba95547f34ce924a53c06. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Modality":"Unimodal","Model":"Wav2Vec 2.0 (LSTM)","Features":"Audio"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.