Emotion Detection using Voice Analysis Utilising EfficientNet and BiLSTM
Bibliographic record
Abstract
Emotion recognition has become a crucial aspect of human-computer interaction, addressing the need for intelligent systems capable of understanding and responding to human emotions. This paper presents a voice-based emotion analysis model that integrates EfficientNet for feature extraction and Bidirectional Long Short-Term Memory (BiLSTM) for sequential processing. The model is trained on the Toronto Emotional Speech Set (TESS) dataset, leveraging acoustic features such as chroma features, spectral contrast, zero-crossing rate, and Mel-Frequency Cepstral Coefficients (MFCCs) to enhance feature representation. Standardization is applied before training to improve model performance. The proposed approach is trained for 50 epochs with a batch size of 32, achieving an accuracy of 82.50% and a macro-average F1 score of 82.52%. While the model demonstrates promising results, challenges remain, including dataset limitations, speaker variability, and the need for improved real-time processing. Future work will focus on addressing these challenges through model enhancements, dataset diversification, and the integration of multimodal emotion recognition techniques to further improve classification accuracy and generalization. The proposed voice-based emotion recognition model enhances real-time human-computer interaction by enabling systems to accurately interpret and respond to users' emotional states. This improvement has practical applications in areas such as virtual assistants, mental health monitoring, customer service, and personalized user experiences.
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 | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.003 | 0.001 |
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".