active learning 2batch selection 1bioacoustics 1budgeted labeling 1disagreement weighting 1facility location 1frame-level audio classification 1gradient embeddings 1long-tailed distribution 1sound event detection 1
From the 2 of 3 linked papers with an AI index.
3 papers
eess.AS2026
Cover First, Disagree Softly: Rethinking Mismatch-First Active Learning for Frame-Level Audio Classification
Shiqi Zhang, Tuomas Virtanen
The paper studies active learning for frame‑level sound event detection and shows that the common mismatch‑first farthest‑traversal strategy performs poorly under limited labeling…
eess.AS2026
Greedy Volume Maximization of Gradient Embeddings for Long-Tailed Frame-Level Bioacoustic Active Learning
Shiqi Zhang, Marius FaiÃ, Ariana Strandburg-Peshkin +1
The paper introduces BADGE‑Greedy‑DPP, a deterministic batch selection method that greedily maximizes the volume of gradient embeddings to improve active learning for sparse, long‑…
eess.AS2026
Mixture-Constrained Max Pooling Improves Separation-Based Bird Species Classification
Yuzhu Wang, Kalle Lahtinen, Patrik Lauha +4
Bird species classification from field recordings remains challenging due to overlapping vocalizations and incomplete species labels. We study source separation as a preprocessing…