Speech Emotions Detection and Classification based on Speech Features using Deep Neural Network
Bibliographic record
Abstract
Automatic speech emotion recognition has received significant research attention in the domain of human-computer interaction during the past ten years. However, the current recognition accuracy has to be improved due to a lack of research on the fundamental temporal relationship of the speech waveform. Usually, speech includes calm, happy, sad, angry, fearful, surprise, and disgust emotions. Detecting and classifying has become a current research challenge. We developed a model to detect the emotion present in the speech data using Long-Short Term Memory(LSTM). The input audio is preprocessed through various techniques such as normalization, trimming, padding and noise reduction. The audio is then fed into the LSTM model to extract the features and detect the emotions. Feature extraction consists of Energy (Root Mean Square), Zero Crossing Rate and Mel Frequency Cepstral Coefficients (MFCCs). Ryerson Audio-Visual Database of Emotional Speech and Song (RAVDESS) and Toronto emotional speech set (TESS) combined datasets are used for testing and training the LSTM model. RAVDESS contains 7356 audio files including those of 24 proficient actors (12 male and 12 female). TESS contains 2800 audio files and emotions set on 200 target words spoken by two artists (aged between 26 to 64 years) and consists of eight emotions such as happiness, anger, fear, disgust, surprise, pleasant, sadness and neutral. Both datasets combined were used to train and test the LSTM model and eight emotions are detected which produces satisfactory outcomes. We obtained validation set and test set accuracy of 94.23% and 97.47% respectively for our system. The outcomes are equated with the state-of-the-art methods which were found to be better. The proposed work can be improved by experimenting with other deep learning techniques such as Recurrent Neural Network (RNN).
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 | 0.001 | 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) | 0.002 | 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".