BSER: A Learning Framework for Bangla Speech Emotion Recognition
Bibliographic record
Abstract
Human Computer Interaction (HCI) relies on accurate speech emotion identification. Speech Emotion Recognition (SER) analyzes voice signals to classify emotions. English based Speech Emotion Recognition (SER) has been extensively studied, while Bangla SER has not. The study integrates a one-dimensional convolution neural network with a long short-term memory (LSTM) architecture into a fully linked network for SER. Speech categorization requires feature inclusion, which this method achieves. We included Additive White Gaussian Noise (AWGN), signal elongation, and pitch alteration to improve dataset dependability. Mel-frequency cepstral coefficients (MFCC), Mel-Spectrogram, Zero Crossing Rate (ZCR), chromagram, and Root Mean Square Error are analyzed in this study. One-dimensional convolutional neural network blocks extract local information, while LSTM layers catch global trends in our model. Training and testing loss curves, confusion matrix, recall, precision, F1-score, and accuracy are used to evaluate the model. We assessed using two cutting-edge datasets, the SUST Bangla Emotional Speech Corpus (SUBESCO) and the Ryerson Audio-Visual Database of Emotional Speech and Song (RAVDESS). Experimental results show that the suggested BSER model is more resilient than baseline models on both datasets. BSER improves research in this sector and shows that our hybrid model can detect and classify emotions in voice inputs.
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 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.001 | 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) | 0.008 | 0.005 |
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".