Research map / Label-efficient and robust learning / Recognition and long-tail
PASA: Progressive-Adaptive Spectral Augmentation for Automated Auscultation in Data-Scarce Environments
Ying Wang, Guoheng Huang, Xueyuan Gong, Xinxin Wang, Xiaochen Yuan
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| replace fixed augmentation with batch-level RL augmentation (ablation) | EfficientNet-B4 · SPRSound 2022 T2-2 | Score (%): 62.80 -> 63.62 | Improved |
| add fixed augmentation to baseline (ablation) | EfficientNet-B4 · SPRSound 2022 T1-2 | Score (%): 66.18 -> 68.80 | Improved |
| add fixed augmentation to baseline (ablation) | EfficientNet-B4 · SPRSound 2022 T2-1 | Score (%): 73.99 -> 73.81 | No clear effect |
| use hybrid batch-sample (HBS) instead of sample-only RL augmentation (ablation) | PASA · SPRSound 2022 T1-2 | Score (%): 69.35 -> 70.39 | Improved |
| use HBS instead of batch-only RL augmentation (ablation) | PASA · SPRSound 2022 T2-1 | Score (%): 74.11 -> 75.71 | Improved |
| use HBS instead of sample-only augmentation on easy task (ablation) | PASA · SPRSound 2022 T1-1 | Score (%): 90.58 -> 90.60 | No clear effect |
| add confidence gap thresholds to difficulty classification (ablation) | PASA HBS · SPRSound 2022 T2-2 | Score (%): 63.85 -> 64.08 | Improved |
| add accuracy thresholds to difficulty classification (ablation) | PASA HBS · SPRSound 2022 T2-2 | Score (%): 64.08 -> 64.29 | Improved |
| raise confidence thresholds from (0.8, 0.5) to (0.9, 0.6) (ablation) | PASA HBS · SPRSound 2022 T2-2 | Score (%): 63.85 -> 63.42 | Hurt |
| increase temperatures from (1.0, 0.3) to (1.5, 0.45) (ablation) | PASA HBS · SPRSound 2022 T2-2 | Score (%): 64.29 -> 63.87 | Hurt |
| use PASA instead of recent SOTA on multi-class task | PASA · SPRSound 2022 T1-2 | Score (%): 84.03 -> 70.39 | Hurt |
| use PASA adaptive augmentation on SPRSound 2023 | EfficientNet-B4 · SPRSound 2023 T1-1 | Score (%): 81.96 -> 84.56 | Improved |
Closest papers by meaning
- Resp-Agent: An Agent-Based System for Multimodal Respiratory Sound Generation and Disease Diagnosis (ICLR 2026 · similarity 0.84)
- Towards Open Respiratory Acoustic Foundation Models: Pretraining and Benchmarking (NeurIPS 2024 · similarity 0.84)
- Fourier-basis Functions to Bridge Augmentation Gap: Rethinking Frequency Augmentation in Image Classification (CVPR 2024 · similarity 0.80)
- E-BATS: Efficient Backpropagation-Free Test-Time Adaptation for Speech Foundation Models (NeurIPS 2025 · similarity 0.79)
- Non-Asymptotic Analysis Of Data Augmentation For Precision Matrix Estimation (NeurIPS 2025 · similarity 0.79)
- Noisy Test-Time Adaptation in Vision-Language Models (ICLR 2025 · similarity 0.79)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
