Karren Yang

Byeongjoo Ahn, Karren Yang, Brian Hamilton et al.

We investigate the benefit of combining blind audio recordings with 3D scene information for novel-view acoustic synthesis. Given audio recordings from 2-4 microphones and the 3D geometry and material of a scene containing multiple unknown sound sources, we estimate the sound anywhere in the scene. We identify the main challenges of novel-view acoustic synthesis as sound source localization, separation, and dereverberation. While naively training an end-to-end network fails to produce high-quality results, we show that incorporating room impulse responses (RIRs) derived from 3D reconstructed rooms enables the same network to jointly tackle these tasks. Our method outperforms existing methods designed for the individual tasks, demonstrating its effectiveness at utilizing 3D visual information. In a simulated study on the Matterport3D-NVAS dataset, our model achieves near-perfect accuracy on source localization, a PSNR of 26.44dB and a SDR of 14.23dB for source separation and dereverberation, resulting in a PSNR of 25.55 dB and a SDR of 14.20 dB on novel-view acoustic synthesis. We release our code and model on our project website at https://github.com/apple/ml-nvas3d. Please wear headphones when listening to the results.

Karren Yang, Ting-Yao Hu, Jen-Hao Rick Chang et al.

Adapting generic speech recognition models to specific individuals is a challenging problem due to the scarcity of personalized data. Recent works have proposed boosting the amount of training data using personalized text-to-speech synthesis. Here, we ask two fundamental questions about this strategy: when is synthetic data effective for personalization, and why is it effective in those cases? To address the first question, we adapt a state-of-the-art automatic speech recognition (ASR) model to target speakers from four benchmark datasets representative of different speaker types. We show that ASR personalization with synthetic data is effective in all cases, but particularly when (i) the target speaker is underrepresented in the global data, and (ii) the capacity of the global model is limited. To address the second question of why personalized synthetic data is effective, we use controllable speech synthesis to generate speech with varied styles and content. Surprisingly, we find that the text content of the synthetic data, rather than style, is important for speaker adaptation. These results lead us to propose a data selection strategy for ASR personalization based on speech content.

Karren D. Yang, Anurag Ranjan, Jen-Hao Rick Chang et al.

We consider the task of animating 3D facial geometry from speech signal. Existing works are primarily deterministic, focusing on learning a one-to-one mapping from speech signal to 3D face meshes on small datasets with limited speakers. While these models can achieve high-quality lip articulation for speakers in the training set, they are unable to capture the full and diverse distribution of 3D facial motions that accompany speech in the real world. Importantly, the relationship between speech and facial motion is one-to-many, containing both inter-speaker and intra-speaker variations and necessitating a probabilistic approach. In this paper, we identify and address key challenges that have so far limited the development of probabilistic models: lack of datasets and metrics that are suitable for training and evaluating them, as well as the difficulty of designing a model that generates diverse results while remaining faithful to a strong conditioning signal as speech. We first propose large-scale benchmark datasets and metrics suitable for probabilistic modeling. Then, we demonstrate a probabilistic model that achieves both diversity and fidelity to speech, outperforming other methods across the proposed benchmarks. Finally, we showcase useful applications of probabilistic models trained on these large-scale datasets: we can generate diverse speech-driven 3D facial motion that matches unseen speaker styles extracted from reference clips; and our synthetic meshes can be used to improve the performance of downstream audio-visual models.

Karren Yang, Wan-Yi Lin, Manash Barman et al.

Beyond achieving high performance across many vision tasks, multimodal models are expected to be robust to single-source faults due to the availability of redundant information between modalities. In this paper, we investigate the robustness of multimodal neural networks against worst-case (i.e., adversarial) perturbations on a single modality. We first show that standard multimodal fusion models are vulnerable to single-source adversaries: an attack on any single modality can overcome the correct information from multiple unperturbed modalities and cause the model to fail. This surprising vulnerability holds across diverse multimodal tasks and necessitates a solution. Motivated by this finding, we propose an adversarially robust fusion strategy that trains the model to compare information coming from all the input sources, detect inconsistencies in the perturbed modality compared to the other modalities, and only allow information from the unperturbed modalities to pass through. Our approach significantly improves on state-of-the-art methods in single-source robustness, achieving gains of 7.8-25.2% on action recognition, 19.7-48.2% on object detection, and 1.6-6.7% on sentiment analysis, without degrading performance on unperturbed (i.e., clean) data.

Hsuan Su, Ting-Yao Hu, Hema Swetha Koppula et al.

While Automatic Speech Recognition (ASR) systems are widely used in many real-world applications, they often do not generalize well to new domains and need to be finetuned on data from these domains. However, target-domain data usually are not readily available in many scenarios. In this paper, we propose a new strategy for adapting ASR models to new target domains without any text or speech from those domains. To accomplish this, we propose a novel data synthesis pipeline that uses a Large Language Model (LLM) to generate a target domain text corpus, and a state-of-the-art controllable speech synthesis model to generate the corresponding speech. We propose a simple yet effective in-context instruction finetuning strategy to increase the effectiveness of LLM in generating text corpora for new domains. Experiments on the SLURP dataset show that the proposed method achieves an average relative word error rate improvement of $28\%$ on unseen target domains without any performance drop in source domains.

Sanjoy Chowdhury, Karren D. Yang, Xudong Liu et al.

Recent multimodal large language models (MLLMs) such as GPT-4o and Qwen3-Omni show strong perception but struggle in multi-speaker, dialogue-centric settings that demand agentic reasoning tracking who speaks, maintaining roles, and grounding events across time. These scenarios are central to multimodal audio-video understanding, where models must jointly reason over audio and visual streams in applications such as conversational video assistants and meeting analytics. We introduce AMUSE, a benchmark designed around tasks that are inherently agentic, requiring models to decompose complex audio-visual interactions into planning, grounding, and reflection steps. It evaluates MLLMs across three modes zero-shot, guided, and agentic and six task families, including spatio-temporal speaker grounding and multimodal dialogue summarization. Across all modes, current models exhibit weak multi-speaker reasoning and inconsistent behavior under both non-agentic and agentic evaluation. Motivated by the inherently agentic nature of these tasks and recent advances in LLM agents, we propose RAFT, a data-efficient agentic alignment framework that integrates reward optimization with intrinsic multimodal self-evaluation as reward and selective parameter adaptation for data and parameter efficient updates. Using RAFT, we achieve up to 39.52\% relative improvement in accuracy on our benchmark. Together, AMUSE and RAFT provide a practical platform for examining agentic reasoning in multimodal models and improving their capabilities.

Karren Yang, Dejan Markovic, Steven Krenn et al.

Since facial actions such as lip movements contain significant information about speech content, it is not surprising that audio-visual speech enhancement methods are more accurate than their audio-only counterparts. Yet, state-of-the-art approaches still struggle to generate clean, realistic speech without noise artifacts and unnatural distortions in challenging acoustic environments. In this paper, we propose a novel audio-visual speech enhancement framework for high-fidelity telecommunications in AR/VR. Our approach leverages audio-visual speech cues to generate the codes of a neural speech codec, enabling efficient synthesis of clean, realistic speech from noisy signals. Given the importance of speaker-specific cues in speech, we focus on developing personalized models that work well for individual speakers. We demonstrate the efficacy of our approach on a new audio-visual speech dataset collected in an unconstrained, large vocabulary setting, as well as existing audio-visual datasets, outperforming speech enhancement baselines on both quantitative metrics and human evaluation studies. Please see the supplemental video for qualitative results at https://github.com/facebookresearch/facestar/releases/download/paper_materials/video.mp4.

Chien-Yu Lin, Qichen Fu, Thomas Merth et al.

Super-resolution (SR) techniques have recently been proposed to upscale the outputs of neural radiance fields (NeRF) and generate high-quality images with enhanced inference speeds. However, existing NeRF+SR methods increase training overhead by using extra input features, loss functions, and/or expensive training procedures such as knowledge distillation. In this paper, we aim to leverage SR for efficiency gains without costly training or architectural changes. Specifically, we build a simple NeRF+SR pipeline that directly combines existing modules, and we propose a lightweight augmentation technique, random patch sampling, for training. Compared to existing NeRF+SR methods, our pipeline mitigates the SR computing overhead and can be trained up to 23x faster, making it feasible to run on consumer devices such as the Apple MacBook. Experiments show our pipeline can upscale NeRF outputs by 2-4x while maintaining high quality, increasing inference speeds by up to 18x on an NVIDIA V100 GPU and 12.8x on an M1 Pro chip. We conclude that SR can be a simple but effective technique for improving the efficiency of NeRF models for consumer devices.

Akshita Gupta, Tatiana Likhomanenko, Karren Dai Yang et al.

The rapid progress of foundation models and large language models (LLMs) has fueled significantly improvement in the capabilities of machine learning systems that benefit from mutlimodal input data. However, existing multimodal models are predominantly built on top of pre-trained LLMs, which can limit accurate modeling of temporal dependencies across other modalities and thus limit the model's ability to jointly process and leverage multimodal inputs. To specifically investigate the alignment of text, video, and speech modalities in LLM-style (decoder-only) models, we consider a simplified multimodal generation task, Video-Text to Speech (VTTS): speech generation conditioned on both its corresponding text and video of talking people. The ultimate goal is to generate speech that not only follows the text but also aligns temporally with the video and is consistent with the facial expressions. In this paper, we first introduce Visatronic, a unified multimodal decoder-only transformer model that adopts an LLM-style architecture to embed visual, textual, and speech inputs into a shared subspace, treating all modalities as temporally aligned token streams. Next, we carefully explore different token mixing strategies to understand the best way to propagate information from the steps where video and text conditioning is input to the steps where the audio is generated. We extensively evaluate Visatronic on the challenging VoxCeleb2 dataset and demonstrate zero-shot generalization to LRS3, where Visatronic, trained on VoxCeleb2, achieves a 4.5% WER, outperforming prior SOTA methods trained only on LRS3, which report a 21.4% WER. Additionally, we propose a new objective metric, TimeSync, specifically designed to measure phoneme-level temporal alignment between generated and reference speech, further ensuring synchronization quality. Demo: https://apple.github.io/visatronic-demo/

Max Müller-Eberstein, Dianna Yee, Karren Yang et al.

Parameter-efficient fine-tuning (PEFT) for personalizing automatic speech recognition (ASR) has recently shown promise for adapting general population models to atypical speech. However, these approaches assume a priori knowledge of the atypical speech disorder being adapted for -- the diagnosis of which requires expert knowledge that is not always available. Even given this knowledge, data scarcity and high inter/intra-speaker variability further limit the effectiveness of traditional fine-tuning. To circumvent these challenges, we first identify the minimal set of model parameters required for ASR adaptation. Our analysis of each individual parameter's effect on adaptation performance allows us to reduce Word Error Rate (WER) by half while adapting 0.03% of all weights. Alleviating the need for cohort-specific models, we next propose the novel use of a meta-learned hypernetwork to generate highly individualized, utterance-level adaptations on-the-fly for a diverse set of atypical speech characteristics. Evaluating adaptation at the global, cohort and individual-level, we show that hypernetworks generalize better to out-of-distribution speakers, while maintaining an overall relative WER reduction of 75.2% using 0.1% of the full parameter budget.

Neha Prasad, Karren Yang, Caroline Uhler

In this paper, we present Super-OT, a novel approach to computational lineage tracing that combines a supervised learning framework with optimal transport based on Generative Adversarial Networks (GANs). Unlike previous approaches to lineage tracing, Super-OT has the flexibility to integrate paired data. We benchmark Super-OT based on single-cell RNA-seq data against Waddington-OT, a popular approach for lineage tracing that also employs optimal transport. We show that Super-OT achieves gains over Waddington-OT in predicting the class outcome of cells during differentiation, since it allows the integration of additional information during training.

Karren Yang, Samuel Goldman, Wengong Jin et al.

In this paper, we aim to synthesize cell microscopy images under different molecular interventions, motivated by practical applications to drug development. Building on the recent success of graph neural networks for learning molecular embeddings and flow-based models for image generation, we propose Mol2Image: a flow-based generative model for molecule to cell image synthesis. To generate cell features at different resolutions and scale to high-resolution images, we develop a novel multi-scale flow architecture based on a Haar wavelet image pyramid. To maximize the mutual information between the generated images and the molecular interventions, we devise a training strategy based on contrastive learning. To evaluate our model, we propose a new set of metrics for biological image generation that are robust, interpretable, and relevant to practitioners. We show quantitatively that our method learns a meaningful embedding of the molecular intervention, which is translated into an image representation reflecting the biological effects of the intervention.

Karren Yang, Bryan Russell, Justin Salamon

Self-supervised audio-visual learning aims to capture useful representations of video by leveraging correspondences between visual and audio inputs. Existing approaches have focused primarily on matching semantic information between the sensory streams. We propose a novel self-supervised task to leverage an orthogonal principle: matching spatial information in the audio stream to the positions of sound sources in the visual stream. Our approach is simple yet effective. We train a model to determine whether the left and right audio channels have been flipped, forcing it to reason about spatial localization across the visual and audio streams. To train and evaluate our method, we introduce a large-scale video dataset, YouTube-ASMR-300K, with spatial audio comprising over 900 hours of footage. We demonstrate that understanding spatial correspondence enables models to perform better on three audio-visual tasks, achieving quantitative gains over supervised and self-supervised baselines that do not leverage spatial audio cues. We also show how to extend our self-supervised approach to 360 degree videos with ambisonic audio.

Karren D. Yang, Caroline Uhler

Multi-domain translation seeks to learn a probabilistic coupling between marginal distributions that reflects the correspondence between different domains. We assume that data from different domains are generated from a shared latent representation based on a structural equation model. Under this assumption, we show that the problem of computing a probabilistic coupling between marginals is equivalent to learning multiple uncoupled autoencoders that embed to a given shared latent distribution. In addition, we propose a new framework and algorithm for multi-domain translation based on learning the shared latent distribution and training autoencoders under distributional constraints. A key practical advantage of our framework is that new autoencoders (i.e., new domains) can be added sequentially to the model without retraining on the other domains, which we demonstrate experimentally on image as well as genomics datasets.

Karren D. Yang, Caroline Uhler

Generative adversarial networks (GANs) are an expressive class of neural generative models with tremendous success in modeling high-dimensional continuous measures. In this paper, we present a scalable method for unbalanced optimal transport (OT) based on the generative-adversarial framework. We formulate unbalanced OT as a problem of simultaneously learning a transport map and a scaling factor that push a source measure to a target measure in a cost-optimal manner. In addition, we propose an algorithm for solving this problem based on stochastic alternating gradient updates, similar in practice to GANs. We also provide theoretical justification for this formulation, showing that it is closely related to an existing static formulation by Liero et al. (2018), and perform numerical experiments demonstrating how this methodology can be applied to population modeling.

Adityanarayanan Radhakrishnan, Karren Yang, Mikhail Belkin et al.

The ability of deep neural networks to generalize well in the overparameterized regime has become a subject of significant research interest. We show that overparameterized autoencoders exhibit memorization, a form of inductive bias that constrains the functions learned through the optimization process to concentrate around the training examples, although the network could in principle represent a much larger function class. In particular, we prove that single-layer fully-connected autoencoders project data onto the (nonlinear) span of the training examples. In addition, we show that deep fully-connected autoencoders learn a map that is locally contractive at the training examples, and hence iterating the autoencoder results in convergence to the training examples. Finally, we prove that depth is necessary and provide empirical evidence that it is also sufficient for memorization in convolutional autoencoders. Understanding this inductive bias may shed light on the generalization properties of overparametrized deep neural networks that are currently unexplained by classical statistical theory.