الملخص
Multi-encoder fusion extends Large Audio-Language Models (LALMs) beyond speech-centric recognition, but selecting encoders via intuition or exhaustive search often introduces redundant representations and inflates an already constrained compute budget. We propose CUES (Correlation-gUided Encoder Selection), a lightweight heuristic that estimates complementarity through task- and category-level Pearson correlations between encoders' performance profiles, scoring a candidate set from single-encoder evaluations alone--without fusion training during selection. Evaluated on the XARES-LLM benchmark with a frozen SmolLM2-135M backbone (LoRA-adapted) via five-fold cross-validation, CUES consistently identifies the same configuration per track from held-out development splits alone, without using test data for selection. For the broad Track~A suite, CUES selects a cross-family trio (Whisper-medium, mHuBERT-147, and Dasheng-base), achieving a 4.3% relative gain over Whisper-medium (0.771 vs. 0.739). For Track~B text generation, it re-anchors on a focused, speech-only pair (mHuBERT-147 and WavLM-base-plus) and actively abstains from adding a divergent encoder, outperforming mHuBERT-147 by 6.3% (0.589 vs. 0.554). Rather than a failure to scale, this divergence is consistent with a diversity--interference trade-off that CUES navigates per track from correlation signals alone: across the evaluated pool, added cross-family diversity tends toward an inverted-U on broad audio tasks but toward steady degradation on text generation, which favors a focused, speech-anchored set.
الكلمات المفتاحية
الموضوع
بيانات النشر
- المجلة
- غير متاح
- وصول مفتوح
- وصول مفتوح أخضر
اقتبس هذه المقالة
APA 7
Liao, P. J., Lee, H. S., Ren, W., Hung, K. H., Lee, H. Y., & Wang, H. M. (2026). Correlation-Guided Encoder Selection for Multi-Encoder Large Audio-Language Models. https://omanscience.com/ar/articles/correlation-guided-encoder-selection-for-multi-encoder-large-audio-language-models
MLA 9
Liao, Pei-Jun, et al. "Correlation-Guided Encoder Selection for Multi-Encoder Large Audio-Language Models." https://omanscience.com/ar/articles/correlation-guided-encoder-selection-for-multi-encoder-large-audio-language-models.
شيكاغو (المؤلف–التاريخ)
Liao, Pei-Jun, Hung-Shin Lee, Wenze Ren, Kuo-Hsuan Hung, Hung-yi Lee, and Hsin-Min Wang. 2026. "Correlation-Guided Encoder Selection for Multi-Encoder Large Audio-Language Models." https://omanscience.com/ar/articles/correlation-guided-encoder-selection-for-multi-encoder-large-audio-language-models.
هارفارد
Liao, P. J., Lee, H. S., Ren, W., Hung, K. H., Lee, H. Y. and Wang, H. M. (2026) 'Correlation-Guided Encoder Selection for Multi-Encoder Large Audio-Language Models', Available at: https://omanscience.com/ar/articles/correlation-guided-encoder-selection-for-multi-encoder-large-audio-language-models.
فانكوفر
Liao PJ, Lee HS, Ren W, Hung KH, Lee HY, Wang HM. Correlation-Guided Encoder Selection for Multi-Encoder Large Audio-Language Models. https://omanscience.com/ar/articles/correlation-guided-encoder-selection-for-multi-encoder-large-audio-language-models
IEEE
P. J. Liao, H. S. Lee, W. Ren, K. H. Hung, H. Y. Lee, and H. M. Wang, "Correlation-Guided Encoder Selection for Multi-Encoder Large Audio-Language Models," https://omanscience.com/ar/articles/correlation-guided-encoder-selection-for-multi-encoder-large-audio-language-models.