Multi-task learning for audio scene source counting and analysis
Bibliographic record
Abstract
Audio source counting is a fundamental task of audio scene analysis related to other audio tasks such as speaker diarization and sound event detection. It is also a relatively unexplored audio task that presents a complex challenge. In particular, source counting performance is poor when the source count range is large, limiting its potential applications. This paper presents a novel approach to improve upon audio source counting through multi-task learning. We present a first of its kind empirical study on the hierarchical nature of audio source counting, introducing the coarse source counting task and a hierarchical multi-task learning framework, in order to better understand and investigate the audio source counting task through several case study scenarios. We perform multi-task learning with a ResNet architecture and demonstrate improvements to audio source counting accuracy by up to a 6% increase from the previous best result on the SARdBScene dataset. We also perform multi-task learning of audio source counting and acoustic scene classification as a step forward for robust audio scene analysis. These experimental results show improvements of up to 6% in source counting accuracy over state-of-the-art baselines, particularly in high source count scenarios. Our findings highlight that multi-task learning not only enhances accuracy, but also improves efficiency by replacing multiple task-specific models with a single robust network.
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 | 0.002 | 0.008 |
| Meta-epidemiology (narrow) | 0.003 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.003 | 0.002 |
| Research integrity | 0.002 | 0.003 |
| Insufficient payload (model declined to judge) | 0.003 | 0.003 |
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".