Recognition
-
Generalizing Multimodal Foundational Models in Context-Driven Class Incremental Learning
2026-10-19 21:13:45Avi Gupta, Nilotpal Sinha, Vishnu Raj, Sambuddha Saha, Pratik Joshi, Koteswar Rao Jerripothula, Tammam TilloAbstract
Class-Incremental Learning (CIL) aims to learn new classes continuously without forgetting the previously learned ones. Recent works in CIL have significantly advanced across various modalities; however, at the intersection of audio-visual modality, this problem remains underexplored. While representations from multimodal foundational models (MFMs) encapsulate rich static priors, our preliminary analysis shows that these representations struggle in incremental settings. Motivated by this, we propose a framework that generalizes frozen MFMs to audio-visual CIL. Specifically, we leverage their dense audio and visual representations and employ an attention-based contextual guidance module, in which the audio representations serve as task-stable contextual guidance for the visual representations, enabling effective class recognition in the incremental setting. To further mitigate catastrophic forgetting and preserve stability across tasks, we employ feature- and logit-level distillation, which are primarily responsible for the reduced forgetting. Extensive evaluations across various audio-visual CIL benchmarks demonstrate that our approach consistently outperforms state-of-the-art methods. Code is available at this https URL.
Abstract (translated)
URL
https://arxiv.org/abs/2610.27298
PDF
-
Preference-Aware Pareto Steering for Multi-Task Speech LLMs
2026-10-19 19:35:51Liuyuan Jiang, Vishal Sunder, Samuel Thomas, Brian Kingsbury, Jeff Kuo, Lisha Chen, George SaonAbstract
Speech large language models (LLMs) perform multiple tasks, including automatic speech recognition (ASR), timestamp prediction (TS), and speech translation (ST). Joint training of speech LLMs creates task trade-offs, yet standard training with fixed loss weights does not generally control the resulting trade-off according to deployment preferences. To address this, we propose Endpoint-Anchored Preference-Aware Sampling Uniformly along the Pareto Front (EA-SURF), an a priori method that adapts scalarization weights so that the resulting task trade-off reflects the desired preference. EA-SURF retains the standard weighted-loss training pipeline, tracking the current model's performance relative to each task-specific reference and adapting the scalarization weights as training proceeds. We theoretically show that EA-SURF can match a desired preference, with the update becoming stationary at the target under checkable regularity conditions. Across six three-objective speech LLM requests, EA-SURF empirically converges toward the requested targets and increases endpoint-inclusive normalized hypervolume by 17.8% over fixed loss weighting.
Abstract (translated)
URL
https://arxiv.org/abs/2610.27200
PDF
-
CARLAR Influence: A Structured Hallucination Risk in Multimodal Cheque Extraction and an Exploratory Compressed-Image Countermeasure
2026-10-19 19:29:12Joong Ho (Joshua), Choi, Aditya Paul, Nidhi SinghAbstract
Multimodal large language models (LLMs) are increasingly used for structured extraction from financial documents, but their behavior under visual ambiguity is poorly understood. We study CARLAR (Courtesy Amount Recognition and Legal Amount Recognition) influence: a model returns the same amount for both cheque fields even though their annotations differ, including cases with a blank annotation encoded as zero. Across ten visual prompting conditions, we evaluate GPT-4.1 and Claude 4.5 Sonnet on 243 primary cheques (124 with two observed unequal amounts) and 598 strict mismatches from two additional datasets. Under courtesy redaction, GPT-4.1 CARLAR reaches 73.7 percent (95 percent bootstrap interval: 68.3 to 79.0; p below 0.001 versus baseline). Model-reported confidence can remain high when predictions agree; expected calibration error reaches 0.93 among CARLAR cases. We also test a compressed-image few-shot intervention that presents examples as one visual table without model fine-tuning. On the 243-case cohort, Claude's Control condition increases paired joint extraction accuracy by 11.5 percentage points (95 percent interval: 6.2 to 16.9; adjusted p = 0.00054), but its strict-124 gain is smaller and not significant after correction. The intervention's effect on predicted-field agreement is model-dependent: it decreases for Claude and increases for GPT-4.1. These results motivate validation methods that account for visual context and dependencies between extracted fields.
Abstract (translated)
URL
https://arxiv.org/abs/2610.27192
PDF
-
Towards Interaction State Recognition in Sensorized Gecko-Inspired Adhesive Interfaces for Orbital Servicing
2026-10-19 18:36:46Eduart Abdullahu, Francesco Grella, Carol MartinezAbstract
Reliable contact establishment remains a critical challenge for autonomous on-orbit servicing (OOS), particularly when interacting with non-cooperative targets. Gecko-inspired dry adhesives provide a lightweight, passive, and residue-free solution for robotic capture in space, yet current systems operate in open loop, without the ability to estimate their own contact or adhesion state during interaction. This paper presents a real-time interaction-awareness framework that enables sensorized gecko-inspired adhesive interfaces to estimate their contact state throughout the adhesion lifecycle using embedded capacitive sensing. Five discrete interaction states (idle, proximity, contact, engaged, and disengaged) are classified using experimentally collected and augmented sensor data, enabling a comparative evaluation of six lightweight machine learning models suitable for embedded deployment. Among the evaluated models, Random Forest achieved the highest offline classification accuracy of 95.65%. For online deployment, combining Random Forest with a rule-based state machine improved accuracy from 78.02% to 93.14%, achieving a macro F1-score of 0.9343 while maintaining a 53 ms inference cycle. These results demonstrate that incorporating physics-consistent state constraints significantly improves robustness during realtime inference, representing a step toward interaction-aware capture in future autonomous orbital servicing missions.
Abstract (translated)
URL
https://arxiv.org/abs/2610.27141
PDF
-
Evaluating Tibetan Speech Recognition Across Dialects and Adaptation Regimes: A Multi-Granularity Analysis
2026-10-19 09:44:36Jacob Moore, Sheng LiAbstract
Tibetan speech recognition is usually evaluated with character or syllable error rates because the writing system does not mark word boundaries. These metrics can rank systems differently and respond differently to adaptation failures. We combine character error rate, syllable error rate, and segmented word error rate in one evaluation framework. We use dictionary-based, neural, and language-model-based segmentation and compare their boundaries with expert annotations. We evaluate five ASR architectures on Lhasa Tibetan, then test zero-shot transfer and target-corpus adaptation on five additional corpus subsets. We also vary LoRA rank and training-data size to examine adaptation failure. The results show that a lower character error does not always lead to a lower syllable or segmented-word error. These ranking differences persist across corpora. Metric correlations also weaken in some unsuccessful adaptation settings. Higher LoRA ranks improve syllable and segmented-word scores but remain worse than full fine-tuning in the tested range. A standardized phonological representation provides another view of spelling errors. These findings support evaluation at several linguistic levels and motivate better segmentation and adaptation methods for low-resource ASR.
Abstract (translated)
URL
https://arxiv.org/abs/2610.27002
PDF
-
Think Before You Transcribe: Context-Steered Reasoning for Hotword-Aware Speech Recognition
2026-10-19 05:09:28Qihao Wang, Hang Su, Fan Cui, Lichun Fan, Jian Luan, Yue HuAbstract
Contextual automatic speech recognition (ASR) improves rare-entity recognition by supplying candidate hotwords, but retrieved candidates may be incomplete or incorrect. A reliable system must therefore exploit a relevant candidate without allowing unsupported alternatives to override the acoustic evidence. We propose a context-steered reasoning framework that generates a short, audio-grounded caption before transcription. This task-specific representation lets the model assess candidate relevance before making a lexical decision. Supervised fine-tuning establishes the structured caption-to-transcription behavior, while targeted trajectory injection strengthens group relative policy optimization when difficult utterances yield only unsuccessful rollouts. Experiments across diverse context conditions show that the proposed method consistently improves both transcription quality and hotword recognition while remaining robust to irrelevant candidates.
Abstract (translated)
URL
https://arxiv.org/abs/2610.26986
PDF
-
How Far Can the Gemma 4 Speech Encoder Go Without Changing Its Weights? Saudi and Gulf Arabic Speech Recognition with a Frozen and an Adapter-Augmented Encoder
2026-10-18 21:06:44Amr AbdelkhalikAbstract
We study the 305M-parameter audio encoder shared by Gemma 4 and EmbeddingGemma 2 (identical except for its final projection) for Saudi and Gulf Arabic speech recognition, without changing any of its weights. A frozen system trains only a 22.6M-parameter CTC reader over a mix of encoder layers; an adapted system adds 3.1M low-rank adapter parameters, a 41.5M-parameter hybrid CTC/attention reader and joint decoding with a sub-word language model. Against Cohere Transcribe Arabic (2.07B parameters), re-run on identical audio, the frozen system trails in word error rate (WER) but is statistically indistinguishable in character error rate (CER) on five of six human-labelled conditions, while running 3.4 times faster on CPU with greedy decoding. The adapted system is statistically indistinguishable in WER on a public Saudi customer-service call (19.20% against 19.27%) and on clean Emirati speech, is significantly better on telephone-coded Emirati speech, has a significantly lower CER on all six conditions of these sets, and beats a 0.6B recogniser fine-tuned on SADA in WER on all six; it still trails Cohere in WER on long-form Saudi speech. On a blind, human-corrected public call-centre sample it ranks 4 of 15 systems, behind three commercial APIs (significantly only one) and ahead of the other self-hostable systems by point estimate. An optimised joint decoder runs it at 139 times real time on one GPU. We also show that references made by correcting machine drafts favour the drafting system. Code and public-set results are released.
Abstract (translated)
URL
https://arxiv.org/abs/2610.28673
PDF
-
FoundPDE: A generative foundation model for data-driven symbolic discovery of partial differential equations
2026-10-18 15:40:42Langchen Liu, Junlin Luo, Min Zhu, Zhuoran Yang, Lu LuAbstract
Discovering governing partial differential equations (PDEs) from observed data is a central inverse problem in computational science and engineering: the goal is not only to predict trajectories, but to identify interpretable symbolic laws that can be analyzed and reused. Existing PDE discovery methods often depend on hand-designed libraries, accurate numerical derivatives, or per-equation symbolic search, making them fragile under sparse observations, noisy data, broad equation families, and unseen symbolic compositions. Here, we introduce FoundPDE, a pretrained data-to-symbol foundation model for PDE discovery. FoundPDE maps discretized spatio-temporal trajectories to symbolic equations with a transformer, uses Fourier and first-order spatial-derivative augmentation to expose complementary trajectory information, refines numerical coefficients, and supports downstream extension with improved generalization through few-shot fine-tuning and hierarchical discovery. On a clean in-domain benchmark spanning 20 PDE families, our model reaches 97.50% type recognition. Under coefficient shifts beyond the training support, type recognition remains 90.53%; under multiplicative Gaussian noise on the model input, it is 94.43% at a 2% noise level but falls to 52.09% at 10%. For unseen composite equations, few-shot fine-tuning reaches over 90% exact recovery on four held-out families, and we further develop a hierarchical discovery algorithm that recovers two-term composites without fine-tuning FoundPDE. By amortizing symbolic discovery across equation families, FoundPDE shifts PDE discovery from per-dataset search toward a reusable pretrained model that extends to new physics with limited additional data, a step toward foundation models that uncover interpretable governing laws directly from scientific data.
Abstract (translated)
URL
https://arxiv.org/abs/2610.26963
PDF
-
VocalGrad: Evaluating Acoustic Perception in Audio Language Models
2026-10-16 16:54:35Koki Ryu, Hitomi YanakaAbstract
Recent large audio language models (LALMs) have extended their capabilities beyond lexical speech understanding to tasks requiring paralinguistic and non-linguistic audio understanding, such as emotion recognition. However, these models still struggle to reliably utilize acoustic information, particularly when it co-occurs with spoken linguistic content. Existing benchmarks do not clearly distinguish acoustic perception from higher-level reasoning, making it difficult to identify the source of these failures. To address this issue, we introduce VocalGrad, a benchmark designed to evaluate LALMs' ability to perceive low-level acoustic variations. VocalGrad formulates this problem as a simple binary classification of temporal change direction, enabling evaluation largely independent of higher-level reasoning or semantic understanding. Our benchmark results reveal a substantial gap in acoustic perception between human annotators and state-of-the-art LALMs: while human annotators achieve high accuracy, LALMs perform at near-chance levels. We further conduct linear probing to identify structural bottlenecks in these models. Our analysis reveals two distinct failure modes: for some categories, task-relevant information is preserved in internal representations but not utilized by the language model head; for others, the information becomes progressively less linearly decodable across model layers. Overall, our results reveal substantial limitations of LALMs in modeling acoustic variation under controlled settings and expose representation differences not observable from text outputs.
Abstract (translated)
URL
https://arxiv.org/abs/2610.23745
PDF
-
Synthetic Image Oversampling for Rare-Class Representation Learning in Deep Facial Phenotyping
2026-10-16 16:43:59Miles Liao, Zhanliang Wang, Hongzhuo Chen, Kai WangAbstract
Recent deep facial phenotyping methods can support the recognition of genetic syndromes, but patient images are often scarce and unevenly distributed across disorders. Synthetic facial images offer a way to supplement the training data without collecting additional patient photographs. Their usefulness, however, depends on whether they improve recognition of real patients: generated images may introduce artifacts or repeat features already represented in the training set. It is therefore necessary to examine both how much synthetic augmentation to use and which disorders benefit. We investigate disease-specific synthetic oversampling using images from the Phenotypic Disease Image Database to augment a subset of the GestaltMatcher Database. We compare increasing augmentation levels in an image-based model, GestaltMatcher-Arc, and a multimodal model, GestaltMML, while keeping the real training and patient-disjoint evaluation partitions fixed. Repeated training runs assess overall and disease-specific accuracy, and an exploratory embedding analysis examines changes in class compactness and separation. Augmentation improves mean recognition accuracy in both training setups, although gains remain modest in GestaltMatcher-Arc and continue across the tested levels in GestaltMML. These aggregate improvements are accompanied by uneven disease-level responses, including reduced accuracy for some disorders. In the embedding analysis, disorders with fewer real training images tend to gain more class separation, but geometric changes do not consistently track classification performance. Together, the findings support synthetic oversampling for limited-data facial phenotyping and emphasize the need to assess its amount and benefit for individual disorders within each training setup.
Abstract (translated)
URL
https://arxiv.org/abs/2610.23730
PDF
-
Recognizing Study Lineage Is Not Enough: Over-Crediting Same-Trial Evidence in LLM-Based Medical Verification
2026-10-16 16:42:10Mukand Krishna, Anoushka Harit, Chenghao Xiao, Zhongtian SunAbstract
Publication count is not study count. A model can identify shared trial lineage correctly in a direct task yet over-credit a second publication from the same trial when aggregating evidence. We evaluate this behavior in medical LLM verification using 14 source-audited clinical families and five model conditions. The families span provenance, clinical applicability, result direction, and composition; the provenance families provide the direct lineage tests. Recognition and aggregation are separate stateless calls, and repeated calls assess within-input stability. Across all three original provenance families, GPT-5.5 identifies the independent update in 10/10 recognition calls per family but over-credits the same-trial report in 5/5 separate aggregation calls per family. Selected Qwen3-4B and Ministral-3-8B conditions show the same cross-task contrast. A study-grouping instruction reduces GPT-5.5 over-credit from 4/5 to 0/5 in each later pure-provenance family. Other model conditions show mixed behavior, and rejected outputs limit several comparisons. The results identify a behavioral gap between elicited lineage recognition and evidence aggregation. The code and data artifacts are available at this https URL.
Abstract (translated)
URL
https://arxiv.org/abs/2610.23721
PDF
-
mmPencil: Toward Writing-Style-Independent In-Air Handwriting Recognition via mmWave Radar and Large Vision-Language Model
2026-10-16 13:08:09Yifan Guo, Zhu Wang, Qian Qin, Yangqian Lei, Qiwen Gan, Zhuo Sun, Chao Chen, Bin Guo, Zhiwen YuAbstract
In-air handwriting recognition based on radio frequency (RF) signals offers a promising solution for human-computer interaction, yet existing methods are often limited by predefined writing planes, constrained writing styles, and small recognition vocabularies. To address these limitations, we propose m^2VLMs, a modality mapping architecture that bridges millimeter-wave (mmWave) radar sensing and large vision-language models (VLMs), and develop mmPencil, a 3D in-air handwriting word recognition system. mmPencil reconstructs 3D handwriting trajectories from mmWave signals using a multi-stage spatial trajectory reconstruction algorithm and maps them into visual representations through a spatial-to-visual mapping algorithm, enabling VLMs to recognize handwriting while reducing sensitivity to writing styles and environmental factors. We collect and release a 3D mmWave handwriting dataset containing 7,664 samples of 200 words from 12 users across 22 writing scenarios. Extensive experiments demonstrate that mmPencil achieves robust word recognition under variations in writing position, range, angle, speed, size, direction, and user movement. It achieves 97.60% accuracy on 200 word classes for seen users and 92.50% zero-shot accuracy on 50 word classes for unseen users, outperforming state-of-the-art baselines.
Abstract (translated)
URL
https://arxiv.org/abs/2610.23419
PDF
-
Flash RNN-T: Memory-Efficient Joint Network and Transducer Loss Computation
Abstract
The recurrent neural network transducer (RNN-T) remains a practical choice for streaming speech recognition, but training requires large intermediate tensors. We present Flash RNN-T, a joint and loss implementation combining specialized kernels with optional tiled recomputation over the full alignment space and vocabulary. Tiled and untiled Flash form the entire measured throughput-memory frontier on medium and long workloads. On H100, untiled Flash achieves 24-39% higher throughput with 26-56% less memory than the fastest baseline on these workloads wherever it fits. Tiling reduces memory further through recomputation. On medium, tiled Flash uses ninefold less memory than that baseline at 16% lower throughput when both use one backward call per batch. On long, it fits where all tested alternatives with that backward schedule run out of memory with 80 GB. These comparisons use each implementation's lowest-memory configuration reaching at least 95% of its peak throughput.
Abstract (translated)
URL
https://arxiv.org/abs/2610.23409
PDF
-
Handwritten Text Recognition Lives in the High-Pixel Variance Subspace
2026-10-16 11:09:27Carlos Garrido-Munoz, Jorge Calvo-ZaragozaAbstract
In self-supervised pretraining for Handwritten Text Recognition (HTR), pixel reconstruction methods outperform contrastive methods, unlike in natural-image classification. We argue that this difference follows from where discriminative signal lies in pixel space: for HTR, it is concentrated in high-variance directions and largely absent from low-variance ones. This predicts that objectives preserving high-variance pixel content will transfer best. We test six SSL methods from three families (pixel-grounded MIM, JEPA, and contrastive) under matched encoder, data, and evaluation protocols on six handwriting benchmarks across five languages. With full labels, pixel-groundrounded SSL achieves the lowest CER on every benchmark and both frozen probes, exposes per-position character information that other families recover only through the readout, and is the only family to benefit from pretraining on real handwriting. Pixel-grounded representations are also more label efficient. Across datasets, encoder alignment with the high-variance pixel subspace predicts CER within every method. With a pretrained LLM decoder, a frozen pixel-grounded encoder is competitive with fully fine-tuned supervised baselines; full fine-tuning achieves the lowest mean CER and ranks first or second on every benchmark. These results show that the value of pixel reconstruction depends on where discriminative signal lies in the input.
Abstract (translated)
URL
https://arxiv.org/abs/2609.35473
PDF
-
Advancing Sentence-Level Dysarthric Audio-Visual Speech Recognition via Test-Time Adaptation
2026-10-16 08:26:55Shiyao Wang, Hui Wang, Jiaming Zhou, Shiwan Zhao, Yong QinAbstract
Dysarthric audio-visual speech recognition (DAVSR) enhances performance by integrating complementary visual cues. However, existing research is largely confined to word-level tasks with limited vocabularies, which severely restricts real-world communication for dysarthric speakers. Furthermore, performance remains heavily dependent on supervised fine-tuning with speaker-specific data; significant variations in age, etiology, and severity cause traditional methods to struggle with unseen speakers. To address these domain shifts and the scarcity of continuous multimodal resources, we first introduce a "source speech search-segmentation" pipeline to construct a structured sentence-level corpus from raw CDSD videos. We then propose the use of unsupervised Test-Time Adaptation (TTA) for sentence-level DAVSR using Sequential-level Generalized Entropy Minimization (SGEM). Our evaluations across the state-of-the-art AV-SUPERB and ViSpeR frameworks demonstrate that TTA consistently alleviates Character Error Rate (CER) degradation for unseen pathological speakers. Paired t-tests verify that these improvements are statistically significant (p < 0.05), highlighting TTA as a promising paradigm for advancing speaker-adaptive assistive speech interfaces.
Abstract (translated)
URL
https://arxiv.org/abs/2610.23117
PDF
-
Known Music, Unknown Response: Leakage-Audited Dynamic Arousal Forecasting With Scheduled Future Music
2026-10-16 07:40:19Bin Wen, Dai-Qiao Zhang, Tien-Ping TanAbstract
Dynamic music emotion recognition usually estimates concurrent or lag-aligned labels from audio that has already been observed. We instead forecast the change in consensus arousal when the upcoming music is scheduled but all later affect labels remain hidden. Evaluation holds out complete songs, encodes each eligible 4-s chunk independently, predicts arousal innovation, and compares the correct future with displaced segments from the same song. On PMEmo and a 1,494-song DEAM evaluation pool excluded from engineering development, a complete past/future MuQ representation raises concordance correlation coefficient (CCC) by 0.389--0.507 over recent affect and remains 0.268--0.382 above a calibrated nonlinear state model. A future-only trajectory contains no past or current audio and still exceeds that state model by 0.150--0.249. The earliest legal future chunk, which ends 5.5--11 s before the target, also remains informative. Correct placement exceeds an overlapping 5-s shift by 0.100--0.209 and the acoustic-nearest eligible same-song segment by 0.078--0.105, with positive song-cluster lower bounds. These results show that both scheduled content and its within-song position carry forecast information under the stated observation rules. The study provides a leakage-controlled reference for known-future music forecasting without treating predictive substitutions as causal interventions.
Abstract (translated)
URL
https://arxiv.org/abs/2610.23063
PDF
-
Recognizing Atari and Assessing Sente: Controlled Evaluation Tasks for Language-Based Go Agents
2026-10-16 05:23:01Taiyi Chen, Iskar Deng, Amandalynne PaulladaAbstract
Domain terminology provides concise descriptions of game states and strategic relationships, but whether game-playing language models can accurately recognize these concepts and relate them to move evaluation remains unclear. We construct a controlled evaluation set centered on atari and its relationship to sente in Go, examining terminology recognition under local board changes and move evaluation under different strategic conditions. We conduct our tests on LoGos, a Go language model, after fine-tuning it with terminology and commentary supervision. The model struggles to recognize changes in the atari relationship and assess its strategic implications. Its terminology predictions are associated with move preferences even when those preferences contradict the reference evaluation. Our controlled tests provide a framework for examining whether language-based Go agents connect domain terminology with precise board relationships and strategic move values.
Abstract (translated)
URL
https://arxiv.org/abs/2610.22953
PDF
-
ASR-Style Transcript Normalization Removes Useful Pragmatic Signal for Dialogue Act Classification
2026-10-16 04:51:37Lucky Onyekwelu-Udoka, Md Shafiqul IslamAbstract
Automatic speech recognition (ASR) pipelines often normalize transcripts by removing filled pauses and punctuation and expanding colloquial forms, potentially discarding cues useful for dialogue act classification. We evaluate nine transcript conditions across six classifiers (TF-IDF logistic regression, RoBERTa, BERT, ALBERT, GPT-2, and BiLSTM) and three conversational corpora (Switchboard, ICSI MRDA, and SLUE-HVB), supplemented by multi-seed experiments, normalization controls, and real-ASR evaluations. Across five random seeds, heavy normalization reduces Macro-F1 in all 15 neural model-corpus combinations under the original preprocessing protocols. Removal using the original filler and backchannel lexicon consistently degrades performance on MRDA and SLUE-HVB but produces mixed effects on Switchboard. Punctuation-aware TF-IDF experiments show that punctuation removal reduces Macro-F1 by 0.036 on Switchboard and 0.078 on MRDA, demonstrating that the original linear baseline's apparent insensitivity resulted from tokenization. Restricting filler removal to um, uh, and erm preserves degradation on MRDA but largely eliminates the SLUE-HVB effect, where the original lexicon removes class-defining backchannel expressions. These findings demonstrate that normalization effects depend on preprocessing definitions, feature extraction, and conversational context, supporting task-dependent rather than indiscriminate transcript normalization.
Abstract (translated)
URL
https://arxiv.org/abs/2610.22934
PDF
-
COMET: Decoupled Distillation, Routing, and Capacity Control for Task-Agnostic Continual Vision--Language Learning
2026-10-16 04:15:49Seungyoon Woo, Keonhee Park, Gunhee KimAbstract
Vision-language models (VLMs) such as CLIP retain zero-shot recognition ability, but their open-world use often requires task-agnostic continual learning: tasks arrive sequentially, past data are unavailable, task identity is unknown, and the backbone cannot be retrained. We argue that this setting fails VLMs through three coupled pressures beyond the stability-plasticity tradeoff: (1) modal-consistency drift, (2) representation-space interference, and (3) per-task capacity shortfall. We propose COMET (Collaborative Mixture-of-Expert Transfer) to resolve these three issues respectively as follows. A frozen teacher provides only a feature reference on a shared image-text pool; Tri-Affinity Distillation preserves image-image, text-text, and image-text geometry; and Student-Aware Cross-Expert Distillation aligns new experts with compatible prior experts so routing errors degrade gracefully. We design COMET to keep capacity and inference separate: a contextual bandit keeps or merges full per-task experts using post-training signals, while a reconstruction router selects sparse experts at test time without teacher semantics. On the X-TAIL benchmark, we show that each proposed component improves continual learning performance. COMET shows that continual vision-language adaptation can preserve multimodal geometry, coordinate accumulated experts, and allocate capacity jointly when feature reference, expert shaping, routing, and capacity control are separated by construction.
Abstract (translated)
URL
https://arxiv.org/abs/2610.22907
PDF
-
Clinically Informed Acoustic-Aware Attribute Modeling for Respiratory Sound Classification
2026-10-16 02:55:07Xinhong Li, Li Xiao, Ziyang Chen, Yuhong Yang, Weiping TuAbstract
Automated respiratory sound classification (RSC) aims to identify normal and adventitious respiratory sounds, such as crackles and wheezes. Most existing methods formulate RSC as a flat four-class classification problem over norma, crackle, wheeze and both. However, unlike conventional classification tasks with semantically independent and mutually exclusive categories, the commonly used four-class RSC labels have a compositional structure and can be represented by two clinically meaningful sound attributes: crackle presence and wheeze presence. Existing flat classifiers do not explicitly exploit this structure, which may increase the learning burden by treating the four labels as independent targets. Meanwhile, crackles and wheezes exhibit distinct temporal characteristics, with sparse transient crackle events and sustained continuous wheeze patterns, making a single temporal aggregation strategy less suitable for modeling different adventitious sound evidence. To address these limitations, we propose an Acoustic-Aware Asymmetric Framework (AAAF) for RSC. AAAF reformulates four-class recognition as two sound-attribute detection tasks and recovers the final category through logical composition during inference. Built upon a pretrained BEATs encoder, AAAF employs branch-specific multi-layer fusion and asymmetric temporal aggregation to capture localized crackle evidence and sustained wheeze patterns. Auxiliary normal--abnormal supervision and a batch-adaptive dynamic loss are further introduced to improve global discrimination and robustness under imbalanced labels. Experiments on ICBHI and SPRSound demonstrate that AAAF improves overall RSC performance and generalizes across datasets, achieving a state-of-the-art four-class Score of 65.30% on ICBHI.
Abstract (translated)
URL
https://arxiv.org/abs/2610.22854
PDF
