{"ID":2858072,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2510.08047","arxiv_id":"2510.08047","title":"Pseudo2Real: Task Arithmetic for Pseudo-Label Correction in Automatic Speech Recognition","abstract":"Robust ASR under domain shift is crucial because real-world systems encounter unseen accents and domains with limited labeled data. Although pseudo-labeling offers a practical workaround, it often introduces systematic, accent-specific errors that filtering fails to fix. We ask: How can we correct these recurring biases without target ground truth? We propose a simple parameter-space correction: in a source domain containing both real and pseudo-labeled data, two ASR models are fine-tuned from the same initialization, one on ground-truth labels and the other on pseudo-labels, and their weight difference forms a correction vector that captures pseudo-label biases. When applied to a pseudo-labeled target model, this vector enhances recognition, achieving up to a 35% relative Word Error Rate (WER) reduction on AfriSpeech-200 across ten African accents with the Whisper tiny model.","short_abstract":"Robust ASR under domain shift is crucial because real-world systems encounter unseen accents and domains with limited labeled data. Although pseudo-labeling offers a practical workaround, it often introduces systematic, accent-specific errors that filtering fails to fix. We ask: How can we correct these recurring biase...","url_abs":"https://arxiv.org/abs/2510.08047","url_pdf":"https://arxiv.org/pdf/2510.08047v2","authors":"[\"Yi-Cheng Lin\",\"Yu-Hsuan Li Liang\",\"Hsuan Su\",\"Tzu-Quan Lin\",\"Shang-Tse Chen\",\"Yun-Nung Chen\",\"Hung-yi Lee\"]","published":"2025-10-09T10:31:47Z","proceeding":"eess.AS","tasks":"[\"eess.AS\",\"cs.CL\"]","methods":"[]","has_code":false}