A Multifaceted Approach Based on Deep Learning Architectures for Dysarthric Speech Recognition
Bibliographic record
Abstract
Dysarthric speech presents significant challenges to effective communication and overall quality of life. Traditional diagnostic methods are often subjective and time-consuming, limiting both the accuracy and efficiency of detecting this condition. In response, this study proposes a novel automated system for dysarthric speech recognition, designed to improve diagnostic precision and timeliness. The proposed system adopts a comprehensive framework that integrates advanced signal processing techniques, robust feature extraction methods, and cutting-edge deep learning architectures. Multiple acoustic features, including Mel-Frequency Cepstral Coefficients “MFCCs” and Power-Normalized Cepstral Coefficients “PNCCs”, are extracted to capture detailed characteristics of the speech signal. Additionally, speech enhancement algorithms such as Minimum Mean Square Error “MMSE” based noise reduction are applied to improve input quality, especially under noisy conditions. At the core of the system lies a deep learning model combining Convolutional Neural Networks “CNNs” with Bidirectional Long Short-Term Memory “BiLSTM” networks. This architecture effectively captures both local and temporal dependencies within speech signals. To further enhance performance, the Connectionist Temporal Classification “CTC” approach is incorporated for word level recognition in dysarthric speech. CTC is particularly beneficial in speech recognition tasks as it enables training sequence to sequence models without requiring strict alignment between input features and output labels. The system's performance has been thoroughly evaluated on the UASpeech dataset, showing notable effectiveness in handling speech with low intelligibility.
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 | 0.000 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 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".