EchoDistill:Alignment Noisy-to-Clean Self-Distillation for Robust Audio LLMs

• URL: http://arxiv.org/abs/2605.23954v1
• Date: Mon, 11 May 2026 06:30:25 GMT
• Title: EchoDistill:Alignment Noisy-to-Clean Self-Distillation for Robust Audio LLMs
• Abstract summary: Echodistill is an alignment-based noisy-to-clean self-distillation framework.<n>It significantly improves semantic reliability and task performance of Audio Large Language Models under complex noise.
• Score: 43.824076052614835
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Audio Large Language Models (ALLMs) are highly vulnerable to real-world noise, which often induces severe semantic drift and hallucinations. Existing robustness methods primarily rely on waveform-level acoustic enhancement, answer-level supervision, or the internal suppression of noise representations. To address these issues, we propose echodistill, an alignment-based noisy-to-clean self-distillation framework. Echodistill leverages a frozen clean-audio teacher to provide semantic references for an inference-time noisy-audio student. Specifically, the student samples candidate responses under noisy conditions to expose its test-time behavior. These trajectories are then optimized via group-relative policy optimization (GRPO), where the token-level consistency with the teacher acts as a reward bonus. By aligning the noisy student's candidate responses with clean semantic evidence, and applying audio-aware reward shaping, our method encourages reasoning trajectories that are both correct and genuinely acoustically grounded. Echodistill significantly improves the semantic reliability and task performance of Audio LLMs under complex noise, without introducing any additional inference costs. Extensive experiments show that: (I) Compared with the strongest baseline, echodistill achieves average improvements of 4.18\%$\uparrow$ in GSR under strong noise. (II) Ablation results on Qwen-Omni further show that echodistill improves over the GRPO-only variant by 3.02\%$\uparrow$ in Acc, 3.89\%$\uparrow$ in Noisy, and 4.53\%$\uparrow$ in GSR on average. Our codes are available at https://anonymous.4open.science/r/echodistill-10DE.

Related papers

• Reinforcement Learning with Evolving Rubrics as Rewards for Audio Reasoning [57.5472764796581]
  We introduce AudioRubrics, a reinforcement learning framework that supervises audio reasoning with self-evolving rewards.<n>AudioRubrics synthesizes per-sample rubrics from the raw waveform and, conditioned on the model's own rollouts, regenerates and reweights criteria per group.<n>Our analysis shows that the gains scale with the capability of the generator and judge, and AudioRubrics converges to a stable reasoning length that avoids both degenerate collapse and growth.
  arXiv  Detail & Related papers  (2026-08-03T19:46:13Z)
• NoiseRater: Meta-Learned Noise Valuation for Diffusion Model Training [81.58321000203712]
  NoiseRater is a meta-learning framework for instance-level noise valuation in diffusion model training.<n>NoiseRater assigns importance scores to individual noise realizations conditioned on data and timestep.<n>Experiments on FFHQ and ImageNet demonstrate that not all noise samples contribute equally.
  arXiv  Detail & Related papers  (2026-05-02T19:43:16Z)
• Pre-train to Gain: Robust Learning Without Clean Labels [1.1582652820340928]
  Training deep networks with noisy labels leads to poor generalization and degraded accuracy.<n>By pre-training a feature extractor backbone without labels, we can train a more noise robust model without requiring a subset with clean labels.<n>Our approach achieves comparable results to ImageNet pre-trained models at low noise levels, while substantially outperforming them under high noise conditions.
  arXiv  Detail & Related papers  (2025-11-25T20:48:07Z)
• Real-Time Speech Enhancement via a Hybrid ViT: A Dual-Input Acoustic-Image Feature Fusion [1.376408511310322]
  Speech quality and intelligibility are significantly degraded in noisy environments.<n>This paper presents a novel transformer-based learning framework to address the single-channel noise suppression problem.
  arXiv  Detail & Related papers  (2025-11-14T19:27:42Z)
• Understanding the Effect of Noise in LLM Training Data with Algorithmic Chains of Thought [0.0]
  We study how noise in chain of thought impacts task performance in highly-controlled setting. We define two types of noise: textitstatic noise, a local form of noise which is applied after the CoT trace is computed, and textitdynamic noise, a global form of noise which propagates errors in the trace as it is computed. We find fine-tuned models are extremely robust to high levels of static noise but struggle significantly more with lower levels of dynamic noise.
  arXiv  Detail & Related papers  (2024-02-06T13:59:56Z)
• Towards Robust Few-shot Point Cloud Semantic Segmentation [57.075074484313]
  Few-shot point cloud semantic segmentation aims to train a model to quickly adapt to new unseen classes with only a handful of support set samples. We propose a Component-level Clean Noise Separation (CCNS) representation learning to learn discriminative feature representations. We also propose a Multi-scale Degree-based Noise Suppression (MDNS) scheme to remove the noisy shots from the support set.
  arXiv  Detail & Related papers  (2023-09-20T11:40:10Z)
• Inference and Denoise: Causal Inference-based Neural Speech Enhancement [83.4641575757706]
  This study addresses the speech enhancement (SE) task within the causal inference paradigm by modeling the noise presence as an intervention. The proposed causal inference-based speech enhancement (CISE) separates clean and noisy frames in an intervened noisy speech using a noise detector and assigns both sets of frames to two mask-based enhancement modules (EMs) to perform noise-conditional SE.
  arXiv  Detail & Related papers  (2022-11-02T15:03:50Z)
• Improving Noise Robustness of Contrastive Speech Representation Learning with Speech Reconstruction [109.44933866397123]
  Noise robustness is essential for deploying automatic speech recognition systems in real-world environments. We employ a noise-robust representation learned by a refined self-supervised framework for noisy speech recognition. We achieve comparable performance to the best supervised approach reported with only 16% of labeled data.
  arXiv  Detail & Related papers  (2021-10-28T20:39:02Z)
• Training Classifiers that are Universally Robust to All Label Noise Levels [91.13870793906968]
  Deep neural networks are prone to overfitting in the presence of label noise. We propose a distillation-based framework that incorporates a new subcategory of Positive-Unlabeled learning. Our framework generally outperforms at medium to high noise levels.
  arXiv  Detail & Related papers  (2021-05-27T13:49:31Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.