Multimodal Emotion Recognition with Cross-Attentive Learning and Feature Fusion
Bibliographic record
Abstract
Multi-Modal Emotion Recognition (MMER) plays a crucial role in enhancing human-computer interaction to interpret and respond to human emotions. While existing methods mostly rely on handcrafted features or simple feature concatenation, we introduce a new approach that refines multimodal fusion through cross-attention, enabling the learning of hierarchical dependencies directly from raw data. This allows for more effective interaction between modalities, improving emotion classification. We propose an MMER framework that integrates audio, text, and video modalities, leveraging deep learning models tailored to each data source. A cross-attention mechanism is employed to fuse information across modalities, ensuring the model focuses on the most salient emotional cues. Additionally, focal loss is used to address class imbalance, enhancing recognition of underrepresented emotional states. Evaluated on the IEMOCAP dataset, the proposed model achieves an average accuracy of 88.31% and an F1-score of 76.43%, outperforming existing state-of-the-art methods. These results demonstrate the system’s robustness in recognizing emotions in complex scenarios, highlighting its potential for real-world applications requiring accurate emotion assessment. The source code is available at: https://github.com/alaaNfissi/Mental-Health-Monitoring-MMER.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.002 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.004 | 0.002 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".