Search papers, labs, and topics across Lattice.
This paper critiques the mismatch-first farthest-traversal (MFFT) strategy for active learning in frame-level audio classification, revealing its limitations in budget-constrained scenarios due to its disregard for segment similarity. The authors introduce a novel approach called mismatch-weighted facility location (MW-FL), which optimally allocates the entire budget by employing a disagreement-weighted coverage objective that penalizes similarity among selected segments. Experimental results demonstrate that MW-FL significantly outperforms MFFT, achieving the best area under the learning curve across multiple datasets by effectively balancing coverage and disagreement in segment selection.
Mismatch-weighted facility location (MW-FL) redefines active learning for audio classification, outperforming traditional methods by prioritizing segment diversity over mere disagreement.
Sound event detection relies on frame-level strong labels whose annotation is expensive. Active learning addresses this problem by selecting the audio segments whose labels help the classifier most. One of the prevailing acquisition strategies for this task, mismatch-first farthest-traversal (MFFT), combines the disagreement between two classifiers and the diversity of the selected segments through hard sequential decisions. It selects whole groups of high-disagreement segments first and spreads only the remaining budget by farthest traversal. On two multi-label datasets we show that this design is blind to the similarity among the selected segments and fails under low budgets, with every mismatch-first variant ending below the plain geometric strategy it builds on. We propose mismatch-weighted facility location (MW-FL), which spends the entire budget through a disagreement-weighted coverage objective that penalizes similarity among the selected segments. The disagreement signal from MFFT is used to obtain the nonnegative weights of this facility-location objective, without introducing hyperparameters. Experiments across two geometric mechanisms with three ways of using disagreement show that coverage of the selected segments is the dominant factor, hard disagreement gating of selection is harmful on both mechanisms, and soft disagreement weighting helps on top of coverage. MW-FL attains the best area under the learning curve on both datasets.
