Do Audio-Visual Large Language Models Really See and Hear?

Audio-Visual Large Language Models (AVLLMs) are emerging as unified interfaces to multimodal perception. We present the first mechanistic interpretability study of AVLLMs, analyzing how audio and visual features evolve and fuse through different layers of an AVLLM to produce the final text outputs. We find that although AVLLMs encode rich audio semantics at intermediate layers, these capabilities largely fail to surface in the final text generation when audio conflicts with vision. Probing analyses show that useful latent audio information is present, but deeper fusion layers disproportionately privilege visual representations that tend to suppress audio cues. We further trace this imbalance to training: the AVLLM's audio behavior strongly matches its vision-language base model, indicating limited additional alignment to audio supervision. Our findings reveal a fundamental modality bias in AVLLMs and provide new mechanistic insights into how multimodal LLMs integrate audio and vision.

SonoWorld: From One Image to a 3D Audio-Visual Scene

Tremendous progress in visual scene generation now turns a single image into an explorable 3D world, yet immersion remains incomplete without sound. We introduce Image2AVScene, the task of generating a 3D audio-visual scene from a single image, and present SonoWorld, the first framework to tackle this challenge. From one image, our pipeline outpaints a 360° panorama, lifts it into a navigable 3D scene, places language-guided sound anchors, and renders ambisonics for point, areal, and ambient sources, yielding spatial audio aligned with scene geometry and semantics. Quantitative evaluations on a newly curated real-world dataset and a controlled user study confirm the effectiveness of our approach. Beyond free-viewpoint audio-visual rendering, we also demonstrate applications to one-shot acoustic learning and audio-visual spatial source separation. Project website: https://humathe.github.io/sonoworld/

AFFORD2ACT: Affordance-Guided Automatic Keypoint Selection for Generalizable and Lightweight Robotic Manipulation

Vision-based robot learning often relies on dense image or point-cloud inputs, which are computationally heavy and entangle irrelevant background features. Existing keypoint-based approaches can focus on manipulation-centric features and be lightweight, but either depend on manual heuristics or task-coupled selection, limiting scalability and semantic understanding. To address this, we propose AFFORD2ACT, an affordance-guided framework that distills a minimal set of semantic 2D keypoints from a text prompt and a single image. AFFORD2ACT follows a three-stage pipeline: affordance filtering, category-level keypoint construction, and transformer-based policy learning with embedded gating to reason about the most relevant keypoints, yielding a compact 38-dimensional state policy that can be trained in 15 minutes, which performs well in real-time without proprioception or dense representations. Across diverse real-world manipulation tasks, AFFORD2ACT consistently improves data efficiency, achieving an 82% success rate on unseen objects, novel categories, backgrounds, and distractors.

Towards Perception-Informed Latent HRTF Representations

Personalized head-related transfer functions (HRTFs) are essential for ensuring a realistic auditory experience over headphones, because they take into account individual anatomical differences that affect listening. Most machine learning approaches to HRTF personalization rely on a learned low-dimensional latent space to generate or select custom HRTFs for a listener. However, these latent representations are typically learned in a manner that optimizes for spectral reconstruction but not for perceptual compatibility, meaning they may not necessarily align with perceptual distance. In this work, we first study whether traditionally learned HRTF representations are well correlated with perceptual relations using auditory-based objective perceptual metrics; we then propose a method for explicitly embedding HRTFs into a perception-informed latent space, leveraging a metric-based loss function and supervision via Metric Multidimensional Scaling (MMDS). Finally, we demonstrate the applicability of these learned representations to the task of HRTF personalization. We suggest that our method has the potential to render personalized spatial audio, leading to an improved listening experience.

ControlTac: Force- and Position-Controlled Tactile Data Augmentation with a Single Reference Image

Vision-based tactile sensing has been widely used in perception, reconstruction, and robotic manipulation. However, collecting large-scale tactile data remains costly due to the localized nature of sensor-object interactions and inconsistencies across sensor instances. Existing approaches to scaling tactile data, such as simulation and free-form tactile generation, often suffer from unrealistic output and poor transferability to downstream tasks. To address this, we propose ControlTac, a two-stage controllable framework that generates realistic tactile images conditioned on a single reference tactile image, contact force, and contact position. With those physical priors as control input, ControlTac generates physically plausible and varied tactile images that can be used for effective data augmentation. Through experiments on three downstream tasks, we demonstrate that ControlTac can effectively augment tactile datasets and lead to consistent gains. Our three real-world experiments further validate the practical utility of our approach. Project page: https://dongyuluo.github.io/controltac.

Learning to Highlight Audio by Watching Movies

Recent years have seen a significant increase in video content creation and consumption. Crafting engaging content requires the careful curation of both visual and audio elements. While visual cue curation, through techniques like optimal viewpoint selection or post-editing, has been central to media production, its natural counterpart, audio, has not undergone equivalent advancements. This often results in a disconnect between visual and acoustic saliency. To bridge this gap, we introduce a novel task: visually-guided acoustic highlighting, which aims to transform audio to deliver appropriate highlighting effects guided by the accompanying video, ultimately creating a more harmonious audio-visual experience. We propose a flexible, transformer-based multimodal framework to solve this task. To train our model, we also introduce a new dataset -- the muddy mix dataset, leveraging the meticulous audio and video crafting found in movies, which provides a form of free supervision. We develop a pseudo-data generation process to simulate poorly mixed audio, mimicking real-world scenarios through a three-step process -- separation, adjustment, and remixing. Our approach consistently outperforms several baselines in both quantitative and subjective evaluation. We also systematically study the impact of different types of contextual guidance and difficulty levels of the dataset. Our project page is here: https://wikichao.github.io/VisAH/.

Differentiable Room Acoustic Rendering with Multi-View Vision Priors

An immersive acoustic experience enabled by spatial audio is just as crucial as the visual aspect in creating realistic virtual environments. However, existing methods for room impulse response estimation rely either on data-demanding learning-based models or computationally expensive physics-based modeling. In this work, we introduce Audio-Visual Differentiable Room Acoustic Rendering (AV-DAR), a framework that leverages visual cues extracted from multi-view images and acoustic beam tracing for physics-based room acoustic rendering. Experiments across six real-world environments from two datasets demonstrate that our multimodal, physics-based approach is efficient, interpretable, and accurate, significantly outperforming a series of prior methods. Notably, on the Real Acoustic Field dataset, AV-DAR achieves comparable performance to models trained on 10 times more data while delivering relative gains ranging from 16.6% to 50.9% when trained at the same scale.

Hearing Anywhere in Any Environment

In mixed reality applications, a realistic acoustic experience in spatial environments is as crucial as the visual experience for achieving true immersion. Despite recent advances in neural approaches for Room Impulse Response (RIR) estimation, most existing methods are limited to the single environment on which they are trained, lacking the ability to generalize to new rooms with different geometries and surface materials. We aim to develop a unified model capable of reconstructing the spatial acoustic experience of any environment with minimum additional measurements. To this end, we present xRIR, a framework for cross-room RIR prediction. The core of our generalizable approach lies in combining a geometric feature extractor, which captures spatial context from panorama depth images, with a RIR encoder that extracts detailed acoustic features from only a few reference RIR samples. To evaluate our method, we introduce ACOUSTICROOMS, a new dataset featuring high-fidelity simulation of over 300,000 RIRs from 260 rooms. Experiments show that our method strongly outperforms a series of baselines. Furthermore, we successfully perform sim-to-real transfer by evaluating our model on four real-world environments, demonstrating the generalizability of our approach and the realism of our dataset.

Aurelia: Test-time Reasoning Distillation in Audio-Visual LLMs

Recent advancements in reasoning optimization have greatly enhanced the performance of large language models (LLMs). However, existing work fails to address the complexities of audio-visual scenarios, underscoring the need for further research. In this paper, we introduce AURELIA, a novel actor-critic based audio-visual (AV) reasoning framework that distills structured, step-by-step reasoning into AVLLMs at test time, improving their ability to process complex multi-modal inputs without additional training or fine-tuning. To further advance AVLLM reasoning skills, we present AVReasonBench, a challenging benchmark comprising 4500 audio-visual questions, each paired with detailed step-by-step reasoning. Our benchmark spans six distinct tasks, including AV-GeoIQ, which evaluates AV reasoning combined with geographical and cultural knowledge. Evaluating 18 AVLLMs on AVReasonBench reveals significant limitations in their multi-modal reasoning capabilities. Using AURELIA, we achieve up to a 100% relative improvement, demonstrating its effectiveness. This performance gain highlights the potential of reasoning-enhanced data generation for advancing AVLLMs in real-world applications. Our code and data will be publicly released at: https: //github.com/schowdhury671/aurelia.

AVTrustBench: Assessing and Enhancing Reliability and Robustness in Audio-Visual LLMs

With the rapid advancement of Multi-modal Large Language Models (MLLMs), several diagnostic benchmarks have recently been developed to assess these models' multi-modal reasoning proficiency. However, these benchmarks are restricted to assessing primarily the visual aspect and do not examine the holistic audio-visual (AV) understanding. Moreover, currently, there are no benchmarks that investigate the capabilities of AVLLMs to calibrate their responses when presented with perturbed inputs. To this end, we introduce Audio-Visual Trustworthiness assessment Benchmark (AVTrustBench), comprising 600K samples spanning over 9 meticulously crafted tasks, evaluating the capabilities of AVLLMs across three distinct dimensions: Adversarial attack, Compositional reasoning, and Modality-specific dependency. Using our benchmark we extensively evaluate 13 state-of-the-art AVLLMs. The findings reveal that the majority of existing models fall significantly short of achieving human-like comprehension, offering valuable insights for future research directions. To alleviate the limitations in the existing approaches, we further propose a robust, model-agnostic calibrated audio-visual preference optimization based training strategy CAVPref, obtaining a gain up to 30.19% across all 9 tasks. We will publicly release our code and benchmark to facilitate future research in this direction.