Breathing pattern classification on ITI dataset (test)
98.8AccuracyATL-TDLM
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| ATL-TDLMComputation Time=Low, Components=AMT + KD-FT + CRL + Hybrid Classifier2026.04 | 98.8 | 98.2 | 97.9 | |
| TL with AMT + Knowledge Distillation-Based Fine-Tuning (KD-FT)Computation Time=Low2026.04 | 97.9 | 97.4 | 97 | |
| TL with Adaptive Multi-Thresholding (AMT)Computation Time=Medium2026.04 | 97.3 | 96.8 | 96.1 | |
| TL with Standard Fine-tuningComputation Time=Medium2026.04 | 96.1 | 95.4 | 94.7 | |
| TL without PreprocessingComputation Time=Medium2026.04 | 94.3 | 93.5 | 92.1 | |
| Traditional CNNComputation Time=High2026.04 | 92.5 | 91.8 | 90.6 |