Guha Balakrishnan

Vishwanath Saragadam, Daniel LeJeune, Jasper Tan et al. · cmu

Implicit neural representations (INRs) have recently advanced numerous vision-related areas. INR performance depends strongly on the choice of the nonlinear activation function employed in its multilayer perceptron (MLP) network. A wide range of nonlinearities have been explored, but, unfortunately, current INRs designed to have high accuracy also suffer from poor robustness (to signal noise, parameter variation, etc.). Inspired by harmonic analysis, we develop a new, highly accurate and robust INR that does not exhibit this tradeoff. Wavelet Implicit neural REpresentation (WIRE) uses a continuous complex Gabor wavelet activation function that is well-known to be optimally concentrated in space-frequency and to have excellent biases for representing images. A wide range of experiments (image denoising, image inpainting, super-resolution, computed tomography reconstruction, image overfitting, and novel view synthesis with neural radiance fields) demonstrate that WIRE defines the new state of the art in INR accuracy, training time, and robustness.

Yiran Sun, Tucker Netherton, Laurence Court et al.

CT scans are the standard-of-care for many clinical ailments, and are needed for treatments like external beam radiotherapy. Unfortunately, CT scanners are rare in low and mid-resource settings due to their costs. Planar X-ray radiography units, in comparison, are far more prevalent, but can only provide limited 2D observations of the 3D anatomy. In this work, we propose a method to generate CT volumes from few (<5) planar X-ray observations using a prior data distribution, and perform the first evaluation of such a reconstruction algorithm for a clinical application: radiotherapy planning. We propose a deep generative model, building on advances in neural implicit representations to synthesize volumetric CT scans from few input planar X-ray images at different angles. To focus the generation task on clinically-relevant features, our model can also leverage anatomical guidance during training (via segmentation masks). We generated 2-field opposed, palliative radiotherapy plans on thoracic CTs reconstructed by our method, and found that isocenter radiation dose on reconstructed scans have <1% error with respect to the dose calculated on clinically acquired CTs using <=4 X-ray views. In addition, our method is better than recent sparse CT reconstruction baselines in terms of standard pixel and structure-level metrics (PSNR, SSIM, Dice score) on the public LIDC lung CT dataset. Code is available at: https://github.com/wanderinrain/Xray2CT.

Dominik Zietlow, Michael Lohaus, Guha Balakrishnan et al.

Algorithmic fairness is frequently motivated in terms of a trade-off in which overall performance is decreased so as to improve performance on disadvantaged groups where the algorithm would otherwise be less accurate. Contrary to this, we find that applying existing fairness approaches to computer vision improve fairness by degrading the performance of classifiers across all groups (with increased degradation on the best performing groups). Extending the bias-variance decomposition for classification to fairness, we theoretically explain why the majority of fairness classifiers designed for low capacity models should not be used in settings involving high-capacity models, a scenario common to computer vision. We corroborate this analysis with extensive experimental support that shows that many of the fairness heuristics used in computer vision also degrade performance on the most disadvantaged groups. Building on these insights, we propose an adaptive augmentation strategy that, uniquely, of all methods tested, improves performance for the disadvantaged groups.

Hao Liang, Pietro Perona, Guha Balakrishnan

We propose an experimental method for measuring bias in face recognition systems. Existing methods to measure bias depend on benchmark datasets that are collected in the wild and annotated for protected (e.g., race, gender) and non-protected (e.g., pose, lighting) attributes. Such observational datasets only permit correlational conclusions, e.g., "Algorithm A's accuracy is different on female and male faces in dataset X.". By contrast, experimental methods manipulate attributes individually and thus permit causal conclusions, e.g., "Algorithm A's accuracy is affected by gender and skin color." Our method is based on generating synthetic faces using a neural face generator, where each attribute of interest is modified independently while leaving all other attributes constant. Human observers crucially provide the ground truth on perceptual identity similarity between synthetic image pairs. We validate our method quantitatively by evaluating race and gender biases of three research-grade face recognition models. Our synthetic pipeline reveals that for these algorithms, accuracy is lower for Black and East Asian population subgroups. Our method can also quantify how perceptual changes in attributes affect face identity distances reported by these models. Our large synthetic dataset, consisting of 48,000 synthetic face image pairs (10,200 unique synthetic faces) and 555,000 human annotations (individual attributes and pairwise identity comparisons) is available to researchers in this important area.

Ahmed Imtiaz Humayun, Randall Balestriero, Guha Balakrishnan et al.

Current Deep Network (DN) visualization and interpretability methods rely heavily on data space visualizations such as scoring which dimensions of the data are responsible for their associated prediction or generating new data features or samples that best match a given DN unit or representation. In this paper, we go one step further by developing the first provably exact method for computing the geometry of a DN's mapping - including its decision boundary - over a specified region of the data space. By leveraging the theory of Continuous Piece-Wise Linear (CPWL) spline DNs, SplineCam exactly computes a DNs geometry without resorting to approximations such as sampling or architecture simplification. SplineCam applies to any DN architecture based on CPWL nonlinearities, including (leaky-)ReLU, absolute value, maxout, and max-pooling and can also be applied to regression DNs such as implicit neural representations. Beyond decision boundary visualization and characterization, SplineCam enables one to compare architectures, measure generalizability and sample from the decision boundary on or off the manifold. Project Website: bit.ly/splinecam.

Yiran Sun, Hana Baroudi, Tucker Netherton et al.

Computed Tomography (CT) scans are the standard-of-care for the visualization and diagnosis of many clinical ailments, and are needed for the treatment planning of external beam radiotherapy. Unfortunately, the availability of CT scanners in low- and mid-resource settings is highly variable. Planar x-ray radiography units, in comparison, are far more prevalent, but can only provide limited 2D observations of the 3D anatomy. In this work we propose DIFR3CT, a 3D latent diffusion model, that can generate a distribution of plausible CT volumes from one or few (<10) planar x-ray observations. DIFR3CT works by fusing 2D features from each x-ray into a joint 3D space, and performing diffusion conditioned on these fused features in a low-dimensional latent space. We conduct extensive experiments demonstrating that DIFR3CT is better than recent sparse CT reconstruction baselines in terms of standard pixel-level (PSNR, SSIM) on both the public LIDC and in-house post-mastectomy CT datasets. We also show that DIFR3CT supports uncertainty quantification via Monte Carlo sampling, which provides an opportunity to measure reconstruction reliability. Finally, we perform a preliminary pilot study evaluating DIFR3CT for automated breast radiotherapy contouring and planning -- and demonstrate promising feasibility. Our code is available at https://github.com/yransun/DIFR3CT.

Hao Liang, Zhixuan Ge, Soumendu Majee et al.

Reconstructing a photorealistic 3D face avatar from a single unconstrained photograph is challenging: feed-forward 3D Gaussian Splatting (3DGS) models degrade on out-of-distribution inputs, while pretrained diffusion models produce high-fidelity images but lack multi-view consistency. We observe that these paradigms are fundamentally complementary: explicit 3D representations guarantee geometric consistency, whereas 2D diffusion priors ensure photorealism. Building on this, we propose SplatShot, a training-free framework that couples these representations directly within the denoising process. Given a base 3DGS face model and a single reference image, we jointly denoise all target views using a per-step 3D feedback loop. At each timestep, we predict clean images from the noisy latents, refit the 3DGS to these multi-view predictions, and back-propagate the photometric discrepancy between the 3DGS re-renderings and 2D predictions into the noise estimate. This steers the sampling trajectory toward strictly 3D-coherent, identity-faithful outputs. Experiments on diverse in-the-wild images demonstrate that SplatShot produces 3D avatars with superior identity preservation, photorealism, and multi-view consistency.

Matt Y. Cheung, Ashok Veeraraghavan, Hanjie Chen et al.

Language model reasoning traces are rarely all-or-nothing; they frequently contain valid intermediate steps before a critical error occurs. Existing uncertainty quantification methods typically certify final answers or entire responses, failing to provide statistical guarantees for the proportion of a sequential trace that can be safely retained. To address this, we introduce CROP (Conformal Reasoning Output Prefixes), a verifier-agnostic calibration procedure for clean-prefix certification. Given any step-level risk proxy, CROP selects a calibrated threshold and returns the longest contiguous prefix whose step risk proxies remain below it, routing the uncertified suffix for downstream review or repair. Assuming exchangeability, CROP rigorously controls the marginal probability that the returned prefix contains an annotated error. Across six process-labeled reasoning datasets, we demonstrate that standard step-level metrics such as AUROC do not fully capture prefix utility, suggesting verifiers should instead be evaluated by certified prefix length. Furthermore, CROP balances over- and under-withholding, improving downstream repair accuracy by preserving valid intermediate reasoning while discarding misleading suffixes. Ultimately, this work positions prefix certification as a rigorous, practical bridge between process supervision, abstention, and repair.

Sepehr Dehdashtian, Ruozhen He, Yi Li et al.

Computer vision systems have witnessed rapid progress over the past two decades due to multiple advances in the field. As these systems are increasingly being deployed in high-stakes real-world applications, there is a dire need to ensure that they do not propagate or amplify any discriminatory tendencies in historical or human-curated data or inadvertently learn biases from spurious correlations. This paper presents a comprehensive survey on fairness that summarizes and sheds light on ongoing trends and successes in the context of computer vision. The topics we discuss include 1) The origin and technical definitions of fairness drawn from the wider fair machine learning literature and adjacent disciplines. 2) Work that sought to discover and analyze biases in computer vision systems. 3) A summary of methods proposed to mitigate bias in computer vision systems in recent years. 4) A comprehensive summary of resources and datasets produced by researchers to measure, analyze, and mitigate bias and enhance fairness. 5) Discussion of the field's success, continuing trends in the context of multimodal foundation and generative models, and gaps that still need to be addressed. The presented characterization should help researchers understand the importance of identifying and mitigating bias in computer vision and the state of the field and identify potential directions for future research.

Mike Wong, Murali Ramanujam, Guha Balakrishnan et al.

Camera orientations (i.e., rotation and zoom) govern the content that a camera captures in a given scene, which in turn heavily influences the accuracy of live video analytics pipelines. However, existing analytics approaches leave this crucial adaptation knob untouched, instead opting to only alter the way that captured images from fixed orientations are encoded, streamed, and analyzed. We present MadEye, a camera-server system that automatically and continually adapts orientations to maximize accuracy for the workload and resource constraints at hand. To realize this using commodity pan-tilt-zoom (PTZ) cameras, MadEye embeds (1) a search algorithm that rapidly explores the massive space of orientations to identify a fruitful subset at each time, and (2) a novel knowledge distillation strategy to efficiently (with only camera resources) select the ones that maximize workload accuracy. Experiments on diverse workloads show that MadEye boosts accuracy by 2.9-25.7% for the same resource usage, or achieves the same accuracy with 2-3.7x lower resource costs.

Kushal Vyas, Ahmed Imtiaz Humayun, Aniket Dashpute et al.

Implicit neural representations (INRs) have demonstrated success in a variety of applications, including inverse problems and neural rendering. An INR is typically trained to capture one signal of interest, resulting in learned neural features that are highly attuned to that signal. Assumed to be less generalizable, we explore the aspect of transferability of such learned neural features for fitting similar signals. We introduce a new INR training framework, STRAINER that learns transferrable features for fitting INRs to new signals from a given distribution, faster and with better reconstruction quality. Owing to the sequential layer-wise affine operations in an INR, we propose to learn transferable representations by sharing initial encoder layers across multiple INRs with independent decoder layers. At test time, the learned encoder representations are transferred as initialization for an otherwise randomly initialized INR. We find STRAINER to yield extremely powerful initialization for fitting images from the same domain and allow for $\approx +10dB$ gain in signal quality early on compared to an untrained INR itself. STRAINER also provides a simple way to encode data-driven priors in INRs. We evaluate STRAINER on multiple in-domain and out-of-domain signal fitting tasks and inverse problems and further provide detailed analysis and discussion on the transferability of STRAINER's features. Our demo can be accessed at https://kushalvyas.github.io/strainer.html .

Hao Liang, Kevin Ni, Guha Balakrishnan

Recent work demonstrates that images from various chest X-ray datasets contain visual features that are strongly correlated with protected demographic attributes like race and gender. This finding raises issues of fairness, since some of these factors may be used by downstream algorithms for clinical predictions. In this work, we propose a framework, using generative adversarial networks (GANs), to visualize what features are most different between X-rays belonging to two demographic subgroups.

Krish Kabra, Kathleen M. Lewis, Guha Balakrishnan

Bias analysis is a crucial step in the process of creating fair datasets for training and evaluating computer vision models. The bottleneck in dataset analysis is annotation, which typically requires: (1) specifying a list of attributes relevant to the dataset domain, and (2) classifying each image-attribute pair. While the second step has made rapid progress in automation, the first has remained human-centered, requiring an experimenter to compile lists of in-domain attributes. However, an experimenter may have limited foresight leading to annotation "blind spots," which in turn can lead to flawed downstream dataset analyses. To combat this, we propose GELDA, a nearly automatic framework that leverages large generative language models (LLMs) to propose and label various attributes for a domain. GELDA takes a user-defined domain caption (e.g., "a photo of a bird," "a photo of a living room") and uses an LLM to hierarchically generate attributes. In addition, GELDA uses the LLM to decide which of a set of vision-language models (VLMs) to use to classify each attribute in images. Results on real datasets show that GELDA can generate accurate and diverse visual attribute suggestions, and uncover biases such as confounding between class labels and background features. Results on synthetic datasets demonstrate that GELDA can be used to evaluate the biases of text-to-image diffusion models and generative adversarial networks. Overall, we show that while GELDA is not accurate enough to replace human annotators, it can serve as a complementary tool to help humans analyze datasets in a cheap, low-effort, and flexible manner.

Hao Liang, Josue Ortega Caro, Vikram Maheshri et al.

Training dataset biases are by far the most scrutinized factors when explaining algorithmic biases of neural networks. In contrast, hyperparameters related to the neural network architecture have largely been ignored even though different network parameterizations are known to induce different implicit biases over learned features. For example, convolutional kernel size is known to affect the frequency content of features learned in CNNs. In this work, we present a causal framework for linking an architectural hyperparameter to out-of-distribution algorithmic bias. Our framework is experimental, in that we train several versions of a network with an intervention to a specific hyperparameter, and measure the resulting causal effect of this choice on performance bias when a particular out-of-distribution image perturbation is applied. In our experiments, we focused on measuring the causal relationship between convolutional kernel size and face analysis classification bias across different subpopulations (race/gender), with respect to high-frequency image details. We show that modifying kernel size, even in one layer of a CNN, changes the frequency content of learned features significantly across data subgroups leading to biased generalization performance even in the presence of a balanced dataset.

Moayed Haji-Ali, Guha Balakrishnan, Vicente Ordonez

Diffusion models have revolutionized image generation in recent years, yet they are still limited to a few sizes and aspect ratios. We propose ElasticDiffusion, a novel training-free decoding method that enables pretrained text-to-image diffusion models to generate images with various sizes. ElasticDiffusion attempts to decouple the generation trajectory of a pretrained model into local and global signals. The local signal controls low-level pixel information and can be estimated on local patches, while the global signal is used to maintain overall structural consistency and is estimated with a reference image. We test our method on CelebA-HQ (faces) and LAION-COCO (objects/indoor/outdoor scenes). Our experiments and qualitative results show superior image coherence quality across aspect ratios compared to MultiDiffusion and the standard decoding strategy of Stable Diffusion. Project page: https://elasticdiffusion.github.io/

Hao Liang, Kevin Ni, Guha Balakrishnan

Recent research demonstrates that deep learning models are capable of precisely extracting bio-information (e.g. race, gender and age) from patients' Chest X-Rays (CXRs). In this paper, we further show that deep learning models are also surprisingly accurate at recognition, i.e., distinguishing CXRs belonging to the same patient from those belonging to different patients. These findings suggest potential privacy considerations that the medical imaging community should consider with the proliferation of large public CXR databases.

Yujin Ham, Mateusz Michalkiewicz, Guha Balakrishnan

3D building reconstruction from imaging data is an important task for many applications ranging from urban planning to reconnaissance. Modern Novel View synthesis (NVS) methods like NeRF and Gaussian Splatting offer powerful techniques for developing 3D models from natural 2D imagery in an unsupervised fashion. These algorithms generally require input training views surrounding the scene of interest, which, in the case of large buildings, is typically not available across all camera elevations. In particular, the most readily available camera viewpoints at scale across most buildings are at near-ground (e.g., with mobile phones) and aerial (drones) elevations. However, due to the significant difference in viewpoint between drone and ground image sets, camera registration - a necessary step for NVS algorithms - fails. In this work we propose a method, DRAGON, that can take drone and ground building imagery as input and produce a 3D NVS model. The key insight of DRAGON is that intermediate elevation imagery may be extrapolated by an NVS algorithm itself in an iterative procedure with perceptual regularization, thereby bridging the visual feature gap between the two elevations and enabling registration. We compiled a semi-synthetic dataset of 9 large building scenes using Google Earth Studio, and quantitatively and qualitatively demonstrate that DRAGON can generate compelling renderings on this dataset compared to baseline strategies.

Delaram Pirhayatifard, Mohammad Taha Toghani, Guha Balakrishnan et al.

In this work, we address the challenge of multi-task image generation with limited data for denoising diffusion probabilistic models (DDPM), a class of generative models that produce high-quality images by reversing a noisy diffusion process. We propose a novel method, SR-DDPM, that leverages representation-based techniques from few-shot learning to effectively learn from fewer samples across different tasks. Our method consists of a core meta architecture with shared parameters, i.e., task-specific layers with exclusive parameters. By exploiting the similarity between diverse data distributions, our method can scale to multiple tasks without compromising the image quality. We evaluate our method on standard image datasets and show that it outperforms both unconditional and conditional DDPM in terms of FID and SSIM metrics.

Yuhao Liu, Sadeer Al-Kindi, Ashok Veeraraghavan et al.

Large-scale pretraining on Earth observation imagery has yielded powerful representations of the natural and built environment. However, most existing geospatial foundation models do not directly model the structured socioeconomic covariates typically stored in tabular form. This modality gap limits their ability to capture the complete total environment, which is critical for reasoning about complex environmental, social, and health-related outcomes. In this work, we propose GeoViSTA (Geospatial Vision-Tabular Transformer), a vision-tabular architecture that learns unified geospatial embeddings from co-registered gridded imagery and tabular data. GeoViSTA utilizes bilateral cross-attention to exchange spatial and semantic information across modalities, guided by a geography-aware attention mechanism that aligns continuous image patches with irregular census-tract tokens. We train GeoViSTA with a self-supervised joint masked-autoencoding objective, forcing it to recover missing image patches and tabular rows using local spatial context and cross-modal cues. Empirically, GeoViSTA's unified embeddings improve linear probing performance on high-impact downstream tasks, outperforming baselines in predicting disease-specific mortality and fire hazard frequency across held-out regions. These results demonstrate that jointly modeling the physical environment alongside structured socioeconomic context yields highly transferable representations for holistic geospatial inference.

Sophia Zorek, Kushal Vyas, Yuhao Liu et al.

Neural fields, also known as implicit neural representations (INRs), offer a powerful framework for modeling continuous geometry, but their effectiveness in high-dimensional scientific settings is limited by slow convergence and scaling challenges. In this study, we extend INR models to handle spatiotemporal and multivariate signals and show how INR features can be transferred across scientific signals to enable efficient and scalable representation across time and ensemble runs in an amortized fashion. Across controlled transformation regimes (e.g., geometric transformations and localized perturbations of synthetic fields) and high-fidelity scientific domains-including turbulent flows, fluid-material impact dynamics, and astrophysical systems-we show that transferable features improve not only signal fidelity but also the accuracy of derived geometric and physical quantities, including density gradients and vorticity. In particular, transferable features reduce iterations to reach target reconstruction quality by up to an order of magnitude, increase early-stage reconstruction quality by multiple dB (with gains exceeding 10 dB in some cases), and consistently improve gradient-based physical accuracy.

Yujin Ham, Junho Kim, Vivek Boominathan et al.

Egocentric "walking tour" videos provide a rich source of image data to develop rich and diverse visual models of environments around the world. However, the significant presence of humans in frames of these videos due to crowds and eye-level camera perspectives mitigates their usefulness in environment modeling applications. We focus on addressing this challenge by developing a generative algorithm that can realistically remove (i.e., inpaint) humans and their associated shadow effects from walking tour videos. Key to our approach is the construction of a rich semi-synthetic dataset of video clip pairs to train this generative model. Each pair in the dataset consists of an environment-only background clip, and a composite clip of walking humans with simulated shadows overlaid on the background. We randomly sourced both foreground and background components from real egocentric walking tour videos around the world to maintain visual diversity. We then used this dataset to fine-tune the state-of-the-art Casper video diffusion model for object and effects inpainting, and demonstrate that the resulting model performs far better than Casper both qualitatively and quantitatively at removing humans from walking tour clips with significant human presence and complex backgrounds. Finally, we show that the resulting generated clips can be used to build successful 3D/4D models of urban locations.

Kushal Vyas, Alper Kayabasi, Daniel Kim et al.

The approximation and convergence properties of implicit neural representations (INRs) are known to be highly sensitive to parameter initialization strategies. While several data-driven initialization methods demonstrate significant improvements over standard random sampling, the reasons for their success -- specifically, whether they encode classical statistical signal priors or more complex features -- remain poorly understood. In this study, we explore this phenomenon through a series of experimental analyses leveraging noise pretraining. We pretrain INRs on diverse noise classes (e.g., Gaussian, Dead Leaves, Spectral) and measure their ability to both fit unseen signals and encode priors for an inverse imaging task (denoising). Our analyses on image and video data reveal a surprising finding: simply pretraining on unstructured noise (Uniform, Gaussian) dramatically improves signal fitting capacity compared to all other baselines. However, unstructured noise also yields poor deep image priors for denoising. In contrast, we also find that noise with the classic $1/|f^α|$ spectral structure of natural images achieves an excellent balance of signal fitting and inverse imaging capabilities, performing on par with the best data-driven initialization methods. This finding enables more efficient INR training in applications lacking sufficient prior domain-specific data. For more details, visit project page at https://kushalvyas.github.io/noisepretraining.html

Kooshan Amini, Yuhao Liu, Jamie Ellen Padgett et al.

Timely and accurate detection of hurricane debris is critical for effective disaster response and community resilience. While post-disaster aerial imagery is readily available, robust debris segmentation solutions applicable across multiple disaster regions remain limited. Developing a generalized solution is challenging due to varying environmental and imaging conditions that alter debris' visual signatures across different regions, further compounded by the scarcity of training data. This study addresses these challenges by fine-tuning pre-trained foundational vision models, achieving robust performance with a relatively small, high-quality dataset. Specifically, this work introduces an open-source dataset comprising approximately 1,200 manually annotated aerial RGB images from Hurricanes Ian, Ida, and Ike. To mitigate human biases and enhance data quality, labels from multiple annotators are strategically aggregated and visual prompt engineering is employed. The resulting fine-tuned model, named fCLIPSeg, achieves a Dice score of 0.70 on data from Hurricane Ida -- a disaster event entirely excluded during training -- with virtually no false positives in debris-free areas. This work presents the first event-agnostic debris segmentation model requiring only standard RGB imagery during deployment, making it well-suited for rapid, large-scale post-disaster impact assessments and recovery planning.

Chandan Singh, Guha Balakrishnan, Pietro Perona

Measuring biases of vision systems with respect to protected attributes like gender and age is critical as these systems gain widespread use in society. However, significant correlations between attributes in benchmark datasets make it difficult to separate algorithmic bias from dataset bias. To mitigate such attribute confounding during bias analysis, we propose a matching approach that selects a subset of images from the full dataset with balanced attribute distributions across protected attributes. Our matching approach first projects real images onto a generative adversarial network (GAN)'s latent space in a manner that preserves semantic attributes. It then finds image matches in this latent space across a chosen protected attribute, yielding a dataset where semantic and perceptual attributes are balanced across the protected attribute. We validate projection and matching strategies with qualitative, quantitative, and human annotation experiments. We demonstrate our work in the context of gender bias in multiple open-source facial-recognition classifiers and find that bias persists after removing key confounders via matching. Code and documentation to reproduce the results here and apply the methods to new data is available at https://github.com/csinva/matching-with-gans .

Amy Zhao, Guha Balakrishnan, Kathleen M. Lewis et al.

We introduce a new video synthesis task: synthesizing time lapse videos depicting how a given painting might have been created. Artists paint using unique combinations of brushes, strokes, and colors. There are often many possible ways to create a given painting. Our goal is to learn to capture this rich range of possibilities. Creating distributions of long-term videos is a challenge for learning-based video synthesis methods. We present a probabilistic model that, given a single image of a completed painting, recurrently synthesizes steps of the painting process. We implement this model as a convolutional neural network, and introduce a novel training scheme to enable learning from a limited dataset of painting time lapses. We demonstrate that this model can be used to sample many time steps, enabling long-term stochastic video synthesis. We evaluate our method on digital and watercolor paintings collected from video websites, and show that human raters find our synthetic videos to be similar to time lapse videos produced by real artists. Our code is available at https://xamyzhao.github.io/timecraft.

Adrian V. Dalca, Guha Balakrishnan, John Guttag et al.

Classical deformable registration techniques achieve impressive results and offer a rigorous theoretical treatment, but are computationally intensive since they solve an optimization problem for each image pair. Recently, learning-based methods have facilitated fast registration by learning spatial deformation functions. However, these approaches use restricted deformation models, require supervised labels, or do not guarantee a diffeomorphic (topology-preserving) registration. Furthermore, learning-based registration tools have not been derived from a probabilistic framework that can offer uncertainty estimates. In this paper, we build a connection between classical and learning-based methods. We present a probabilistic generative model and derive an unsupervised learning-based inference algorithm that uses insights from classical registration methods and makes use of recent developments in convolutional neural networks (CNNs). We demonstrate our method on a 3D brain registration task for both images and anatomical surfaces, and provide extensive empirical analyses. Our principled approach results in state of the art accuracy and very fast runtimes, while providing diffeomorphic guarantees. Our implementation is available at http://voxelmorph.csail.mit.edu.

Amy Zhao, Guha Balakrishnan, Frédo Durand et al.

Image segmentation is an important task in many medical applications. Methods based on convolutional neural networks attain state-of-the-art accuracy; however, they typically rely on supervised training with large labeled datasets. Labeling medical images requires significant expertise and time, and typical hand-tuned approaches for data augmentation fail to capture the complex variations in such images. We present an automated data augmentation method for synthesizing labeled medical images. We demonstrate our method on the task of segmenting magnetic resonance imaging (MRI) brain scans. Our method requires only a single segmented scan, and leverages other unlabeled scans in a semi-supervised approach. We learn a model of transformations from the images, and use the model along with the labeled example to synthesize additional labeled examples. Each transformation is comprised of a spatial deformation field and an intensity change, enabling the synthesis of complex effects such as variations in anatomy and image acquisition procedures. We show that training a supervised segmenter with these new examples provides significant improvements over state-of-the-art methods for one-shot biomedical image segmentation. Our code is available at https://github.com/xamyzhao/brainstorm.

Adrian V. Dalca, Guha Balakrishnan, John Guttag et al.

Traditional deformable registration techniques achieve impressive results and offer a rigorous theoretical treatment, but are computationally intensive since they solve an optimization problem for each image pair. Recently, learning-based methods have facilitated fast registration by learning spatial deformation functions. However, these approaches use restricted deformation models, require supervised labels, or do not guarantee a diffeomorphic (topology-preserving) registration. Furthermore, learning-based registration tools have not been derived from a probabilistic framework that can offer uncertainty estimates. In this paper, we present a probabilistic generative model and derive an unsupervised learning-based inference algorithm that makes use of recent developments in convolutional neural networks (CNNs). We demonstrate our method on a 3D brain registration task, and provide an empirical analysis of the algorithm. Our approach results in state of the art accuracy and very fast runtimes, while providing diffeomorphic guarantees and uncertainty estimates. Our implementation is available online at http://voxelmorph.csail.mit.edu .

Guha Balakrishnan, Amy Zhao, Mert R. Sabuncu et al.

We present a fast learning-based algorithm for deformable, pairwise 3D medical image registration. Current registration methods optimize an objective function independently for each pair of images, which can be time-consuming for large data. We define registration as a parametric function, and optimize its parameters given a set of images from a collection of interest. Given a new pair of scans, we can quickly compute a registration field by directly evaluating the function using the learned parameters. We model this function using a convolutional neural network (CNN), and use a spatial transform layer to reconstruct one image from another while imposing smoothness constraints on the registration field. The proposed method does not require supervised information such as ground truth registration fields or anatomical landmarks. We demonstrate registration accuracy comparable to state-of-the-art 3D image registration, while operating orders of magnitude faster in practice. Our method promises to significantly speed up medical image analysis and processing pipelines, while facilitating novel directions in learning-based registration and its applications. Our code is available at https://github.com/balakg/voxelmorph .

Jeffri Murrugarra-Llerena, Pranav Chitale, Zicheng Liu et al.

Social group detection, or the identification of humans involved in reciprocal interpersonal interactions (e.g., family members, friends, and customers and merchants), is a crucial component of social intelligence needed for agents transacting in the world. The few existing benchmarks for social group detection are limited by low scene diversity and reliance on third-person camera sources (e.g., surveillance footage). Consequently, these benchmarks generally lack real-world evaluation on how groups form and evolve in diverse cultural contexts and unconstrained settings. To address this gap, we introduce EgoGroups, a first-person view dataset that captures social dynamics in cities around the world. EgoGroups spans 65 countries covering low, medium, and high-crowd settings under four weather/time-of-day conditions. We include dense human annotations for person and social groups, along with rich geographic and scene metadata. Using this dataset, we performed an extensive evaluation of state-of-the-art VLM/LLMs and supervised models on their group detection capabilities. We found several interesting findings, including VLMs and LLMs can outperform supervised baselines in a zero-shot setting, while crowd density and cultural regions clearly influence model performance.

Matt Y Cheung, Tucker J Netherton, Laurence E Court et al.

Modern deep learning reconstruction algorithms generate impressively realistic scans from sparse inputs, but can often produce significant inaccuracies. This makes it difficult to provide statistically guaranteed claims about the true state of a subject from scans reconstructed by these algorithms. In this study, we propose a framework for computing provably valid prediction bounds on claims derived from probabilistic black-box image reconstruction algorithms. The key insights behind our framework are to represent reconstructed scans with a derived clinical metric of interest, and to calibrate bounds on the ground truth metric with conformal prediction (CP) using a prior calibration dataset. These bounds convey interpretable feedback about the subject's state, and can also be used to retrieve nearest-neighbor reconstructed scans for visual inspection. We demonstrate the utility of this framework on sparse-view computed tomography (CT) for fat mass quantification and radiotherapy planning tasks. Results show that our framework produces bounds with better semantical interpretation than conventional pixel-based bounding approaches. Furthermore, we can flag dangerous outlier reconstructions that look plausible but have statistically unlikely metric values.

Alper Kayabasi, Anil Kumar Vadathya, Guha Balakrishnan et al.

We propose a new continuous video modeling framework based on implicit neural representations (INRs) called ActINR. At the core of our approach is the observation that INRs can be considered as a learnable dictionary, with the shapes of the basis functions governed by the weights of the INR, and their locations governed by the biases. Given compact non-linear activation functions, we hypothesize that an INR's biases are suitable to capture motion across images, and facilitate compact representations for video sequences. Using these observations, we design ActINR to share INR weights across frames of a video sequence, while using unique biases for each frame. We further model the biases as the output of a separate INR conditioned on time index to promote smoothness. By training the video INR and this bias INR together, we demonstrate unique capabilities, including $10\times$ video slow motion, 4x spatial super resolution along with 2x slow motion, denoising, and video inpainting. ActINR performs remarkably well across numerous video processing tasks (often achieving more than 6dB improvement), setting a new standard for continuous modeling of videos.

Jefferson Hernandez, Ruozhen He, Guha Balakrishnan et al.

We introduce GVIT, a classification framework that abandons conventional pixel or patch grid input representations in favor of a compact set of learnable 2D Gaussians. Each image is encoded as a few hundred Gaussians whose positions, scales, orientations, colors, and opacities are optimized jointly with a ViT classifier trained on top of these representations. We reuse the classifier gradients as constructive guidance, steering the Gaussians toward class-salient regions while a differentiable renderer optimizes an image reconstruction loss. We demonstrate that by 2D Gaussian input representations coupled with our GVIT guidance, using a relatively standard ViT architecture, closely matches the performance of a traditional patch-based ViT, reaching a 76.9% top-1 accuracy on Imagenet-1k using a ViT-B architecture.

Tony Yu, Yanlin Jin, Ashok Veeraraghavan et al.

We present TranSplat, a 3D scene rendering algorithm that enables realistic cross-scene object transfer (from a source to a target scene) based on the Gaussian Splatting framework. Our approach addresses two critical challenges: (1) precise 3D object extraction from the source scene, and (2) faithful relighting of the transferred object in the target scene without explicit material property estimation. TranSplat fits a splatting model to the source scene, using 2D object masks to drive fine-grained 3D segmentation. Following user-guided insertion of the object into the target scene, along with automatic refinement of position and orientation, TranSplat derives per-Gaussian radiance transfer functions via spherical harmonic analysis to adapt the object's appearance to match the target scene's lighting environment. This relighting strategy does not require explicitly estimating physical scene properties such as BRDFs. Evaluated on several synthetic and real-world scenes and objects, TranSplat yields excellent 3D object extractions and relighting performance compared to recent baseline methods and visually convincing cross-scene object transfers. We conclude by discussing the limitations of the approach.

Mateusz Michalkiewicz, Sheena Bai, Mahsa Baktashmotlagh et al.

In this paper, we analyze the viewpoint stability of foundational models - specifically, their sensitivity to changes in viewpoint- and define instability as significant feature variations resulting from minor changes in viewing angle, leading to generalization gaps in 3D reasoning tasks. We investigate nine foundational models, focusing on their responses to viewpoint changes, including the often-overlooked accidental viewpoints where specific camera orientations obscure an object's true 3D structure. Our methodology enables recognizing and classifying out-of-distribution (OOD), accidental, and stable viewpoints using feature representations alone, without accessing the actual images. Our findings indicate that while foundation models consistently encode accidental viewpoints, they vary in their interpretation of OOD viewpoints due to inherent biases, at times leading to object misclassifications based on geometric resemblance. Through quantitative and qualitative evaluations on three downstream tasks - classification, VQA, and 3D reconstruction - we illustrate the impact of viewpoint instability and underscore the importance of feature robustness across diverse viewing conditions.

Kushal Vyas, Ashok Veeraraghavan, Guha Balakrishnan

Implicit neural representations (INRs) have achieved remarkable successes in learning expressive yet compact signal representations. However, they are not naturally amenable to predictive tasks such as segmentation, where they must learn semantic structures over a distribution of signals. In this study, we introduce MetaSeg, a meta-learning framework to train INRs for medical image segmentation. MetaSeg uses an underlying INR that simultaneously predicts per pixel intensity values and class labels. It then uses a meta-learning procedure to find optimal initial parameters for this INR over a training dataset of images and segmentation maps, such that the INR can simply be fine-tuned to fit pixels of an unseen test image, and automatically decode its class labels. We evaluated MetaSeg on 2D and 3D brain MRI segmentation tasks and report Dice scores comparable to commonly used U-Net models, but with $90\%$ fewer parameters. MetaSeg offers a fresh, scalable alternative to traditional resource-heavy architectures such as U-Nets and vision transformers for medical image segmentation. Our project is available at https://kushalvyas.github.io/metaseg.html .

Matt Y. Cheung, Ashok Veeraraghavan, Guha Balakrishnan

In clinical applications, the utility of segmentation models is often based on the accuracy of derived downstream metrics such as organ size, rather than by the pixel-level accuracy of the segmentation masks themselves. Thus, uncertainty quantification for such metrics is crucial for decision-making. Conformal prediction (CP) is a popular framework to derive such principled uncertainty guarantees, but applying CP naively to the final scalar metric is inefficient because it treats the complex, non-linear segmentation-to-metric pipeline as a black box. We introduce COMPASS, a practical framework that generates efficient, metric-based CP intervals for image segmentation models by leveraging the inductive biases of their underlying deep neural networks. COMPASS performs calibration directly in the model's representation space by perturbing intermediate features along low-dimensional subspaces maximally sensitive to the target metric. We prove that COMPASS achieves valid marginal coverage under exchangeability and nestedness assumptions. Empirically, we demonstrate that COMPASS produces significantly tighter intervals than traditional CP baselines on four medical image segmentation tasks for area estimation of skin lesions and anatomical structures. Furthermore, we show that leveraging learned internal features to estimate importance weights allows COMPASS to also recover target coverage under covariate shifts. COMPASS paves the way for practical, metric-based uncertainty quantification for medical image segmentation.

Hao Liang, Zhixuan Ge, Ashish Tiwari et al.

We present FastAvatar, a pose-invariant, feed-forward framework that can generate a 3D Gaussian Splatting (3DGS) model from a single face image from an arbitrary pose in near-instant time (<10ms). FastAvatar uses a novel encoder-decoder neural network design to achieve both fast fitting and identity preservation regardless of input pose. First, FastAvatar constructs a 3DGS face ``template'' model from a training dataset of faces with multi-view captures. Second, FastAvatar encodes the input face image into an identity-specific and pose-invariant latent embedding, and decodes this embedding to predict residuals to the structural and appearance parameters of each Gaussian in the template 3DGS model. By only inferring residuals in a feed-forward fashion, model inference is fast and robust. FastAvatar significantly outperforms existing feed-forward face 3DGS methods (e.g., GAGAvatar) in reconstruction quality, and runs 1000x faster than per-face optimization methods (e.g., FlashAvatar, GaussianAvatars and GASP). In addition, FastAvatar's novel latent space design supports real-time identity interpolation and attribute editing which is not possible with any existing feed-forward 3DGS face generation framework. FastAvatar's combination of excellent reconstruction quality and speed expands the scope of 3DGS for photorealistic avatar applications in consumer and interactive systems.

Mateusz Michalkiewicz, Anekha Sokhal, Tadeusz Michalkiewicz et al.

Monocular 3D reconstruction methods and vision-language models (VLMs) demonstrate impressive results on standard benchmarks, yet their true understanding of geometric properties remains unclear. We introduce GIQ , a comprehensive benchmark specifically designed to evaluate the geometric reasoning capabilities of vision and vision-language foundation models. GIQ comprises synthetic and real-world images of 224 diverse polyhedra - including Platonic, Archimedean, Johnson, and Catalan solids, as well as stellations and compound shapes - covering varying levels of complexity and symmetry. Through systematic experiments involving monocular 3D reconstruction, 3D symmetry detection, mental rotation tests, and zero-shot shape classification tasks, we reveal significant shortcomings in current models. State-of-the-art reconstruction algorithms trained on extensive 3D datasets struggle to reconstruct even basic geometric forms accurately. While foundation models effectively detect specific 3D symmetry elements via linear probing, they falter significantly in tasks requiring detailed geometric differentiation, such as mental rotation. Moreover, advanced vision-language assistants exhibit remarkably low accuracy on complex polyhedra, systematically misinterpreting basic properties like face geometry, convexity, and compound structures. GIQ is publicly available, providing a structured platform to highlight and address critical gaps in geometric intelligence, facilitating future progress in robust, geometry-aware representation learning.

Matt Y. Cheung, Sophia Zorek, Tucker J. Netherton et al.

Diffusion models demonstrate state-of-the-art performance on image generation, and are gaining traction for sparse medical image reconstruction tasks. However, compared to classical reconstruction algorithms relying on simple analytical priors, diffusion models have the dangerous property of producing realistic looking results \emph{even when incorrect}, particularly with few observations. We investigate the utility of diffusion models as priors for image reconstruction by varying the number of observations and comparing their performance to classical priors (sparse and Tikhonov regularization) using pixel-based, structural, and downstream metrics. We make comparisons on low-dose chest wall computed tomography (CT) for fat mass quantification. First, we find that classical priors are superior to diffusion priors when the number of projections is ``sufficient''. Second, we find that diffusion priors can capture a large amount of detail with very few observations, significantly outperforming classical priors. However, they fall short of capturing all details, even with many observations. Finally, we find that the performance of diffusion priors plateau after extremely few ($\approx$10-15) projections. Ultimately, our work highlights potential issues with diffusion-based sparse reconstruction and underscores the importance of further investigation, particularly in high-stakes clinical settings.

Moayed Haji-Ali, Willi Menapace, Aliaksandr Siarohin et al.

The scalability of ambient sound generators is hindered by data scarcity, insufficient caption quality, and limited scalability in model architecture. This work addresses these challenges by advancing both data and model scaling. First, we propose an efficient and scalable dataset collection pipeline tailored for ambient audio generation, resulting in AutoReCap-XL, the largest ambient audio-text dataset with over 47 million clips. To provide high-quality textual annotations, we propose AutoCap, a high-quality automatic audio captioning model. By adopting a Q-Former module and leveraging audio metadata, AutoCap substantially enhances caption quality, reaching a CIDEr score of $83.2$, a $3.2\%$ improvement over previous captioning models. Finally, we propose GenAu, a scalable transformer-based audio generation architecture that we scale up to 1.25B parameters. We demonstrate its benefits from data scaling with synthetic captions as well as model size scaling. When compared to baseline audio generators trained at similar size and data scale, GenAu obtains significant improvements of $4.7\%$ in FAD score, $11.1\%$ in IS, and $13.5\%$ in CLAP score. Our code, model checkpoints, and dataset are publicly available.

Krish Kabra, Guha Balakrishnan

Perceptual metrics, like the Fréchet Inception Distance (FID), are widely used to assess the similarity between synthetically generated and ground truth (real) images. The key idea behind these metrics is to compute errors in a deep feature space that captures perceptually and semantically rich image features. Despite their popularity, the effect that different deep features and their design choices have on a perceptual metric has not been well studied. In this work, we perform a causal analysis linking differences in semantic attributes and distortions between face image distributions to Fréchet distances (FD) using several popular deep feature spaces. A key component of our analysis is the creation of synthetic counterfactual faces using deep face generators. Our experiments show that the FD is heavily influenced by its feature space's training dataset and objective function. For example, FD using features extracted from ImageNet-trained models heavily emphasize hats over regions like the eyes and mouth. Moreover, FD using features from a face gender classifier emphasize hair length more than distances in an identity (recognition) feature space. Finally, we evaluate several popular face generation models across feature spaces and find that StyleGAN2 consistently ranks higher than other face generators, except with respect to identity (recognition) features. This suggests the need for considering multiple feature spaces when evaluating generative models and using feature spaces that are tuned to nuances of the domain of interest.

Vishwanath Saragadam, Jasper Tan, Guha Balakrishnan et al.

We introduce a new neural signal model designed for efficient high-resolution representation of large-scale signals. The key innovation in our multiscale implicit neural representation (MINER) is an internal representation via a Laplacian pyramid, which provides a sparse multiscale decomposition of the signal that captures orthogonal parts of the signal across scales. We leverage the advantages of the Laplacian pyramid by representing small disjoint patches of the pyramid at each scale with a small MLP. This enables the capacity of the network to adaptively increase from coarse to fine scales, and only represent parts of the signal with strong signal energy. The parameters of each MLP are optimized from coarse-to-fine scale which results in faster approximations at coarser scales, thereby ultimately an extremely fast training process. We apply MINER to a range of large-scale signal representation tasks, including gigapixel images and very large point clouds, and demonstrate that it requires fewer than 25% of the parameters, 33% of the memory footprint, and 10% of the computation time of competing techniques such as ACORN to reach the same representation accuracy.

Guha Balakrishnan, Raghudeep Gadde, Aleix Martinez et al.

We present a method for finding paths in a deep generative model's latent space that can maximally vary one set of image features while holding others constant. Crucially, unlike past traversal approaches, ours can manipulate multidimensional features of an image such as facial identity and pixels within a specified region. Our method is principled and conceptually simple: optimal traversal directions are chosen by maximizing differential changes to one feature set such that changes to another set are negligible. We show that this problem is nearly equivalent to one of Rayleigh quotient maximization, and provide a closed-form solution to it based on solving a generalized eigenvalue equation. We use repeated computations of the corresponding optimal directions, which we call Rayleigh EigenDirections (REDs), to generate appropriately curved paths in latent space. We empirically evaluate our method using StyleGAN2 on two image domains: faces and living rooms. We show that our method is capable of controlling various multidimensional features out of the scope of previous latent space traversal methods: face identity, spatial frequency bands, pixels within a region, and the appearance and position of an object. Our work suggests that a wealth of opportunities lies in the local analysis of the geometry and semantics of latent spaces.

Divya Shanmugam, Davis Blalock, Guha Balakrishnan et al.

Test-time augmentation -- the aggregation of predictions across transformed versions of a test input -- is a common practice in image classification. Traditionally, predictions are combined using a simple average. In this paper, we present 1) experimental analyses that shed light on cases in which the simple average is suboptimal and 2) a method to address these shortcomings. A key finding is that even when test-time augmentation produces a net improvement in accuracy, it can change many correct predictions into incorrect predictions. We delve into when and why test-time augmentation changes a prediction from being correct to incorrect and vice versa. Building on these insights, we present a learning-based method for aggregating test-time augmentations. Experiments across a diverse set of models, datasets, and augmentations show that our method delivers consistent improvements over existing approaches.

Guha Balakrishnan, Yuanjun Xiong, Wei Xia et al.

Measuring algorithmic bias is crucial both to assess algorithmic fairness, and to guide the improvement of algorithms. Current methods to measure algorithmic bias in computer vision, which are based on observational datasets, are inadequate for this task because they conflate algorithmic bias with dataset bias. To address this problem we develop an experimental method for measuring algorithmic bias of face analysis algorithms, which manipulates directly the attributes of interest, e.g., gender and skin tone, in order to reveal causal links between attribute variation and performance change. Our proposed method is based on generating synthetic ``transects'' of matched sample images that are designed to differ along specific attributes while leaving other attributes constant. A crucial aspect of our approach is relying on the perception of human observers, both to guide manipulations, and to measure algorithmic bias. Besides allowing the measurement of algorithmic bias, synthetic transects have other advantages with respect to observational datasets: they sample attributes more evenly allowing for more straightforward bias analysis on minority and intersectional groups, they enable prediction of bias in new scenarios, they greatly reduce ethical and legal challenges, and they are economical and fast to obtain, helping make bias testing affordable and widely available. We validate our method by comparing it to a study that employs the traditional observational method for analyzing bias in gender classification algorithms. The two methods reach different conclusions. While the observational method reports gender and skin color biases, the experimental method reveals biases due to gender, hair length, age, and facial hair.

Guha Balakrishnan, Adrian V. Dalca, Amy Zhao et al.

We introduce visual deprojection: the task of recovering an image or video that has been collapsed along a dimension. Projections arise in various contexts, such as long-exposure photography, where a dynamic scene is collapsed in time to produce a motion-blurred image, and corner cameras, where reflected light from a scene is collapsed along a spatial dimension because of an edge occluder to yield a 1D video. Deprojection is ill-posed-- often there are many plausible solutions for a given input. We first propose a probabilistic model capturing the ambiguity of the task. We then present a variational inference strategy using convolutional neural networks as functional approximators. Sampling from the inference network at test time yields plausible candidates from the distribution of original signals that are consistent with a given input projection. We evaluate the method on several datasets for both spatial and temporal deprojection tasks. We first demonstrate the method can recover human gait videos and face images from spatial projections, and then show that it can recover videos of moving digits from dramatically motion-blurred images obtained via temporal projection.

Guha Balakrishnan, Amy Zhao, Mert R. Sabuncu et al.

We present VoxelMorph, a fast learning-based framework for deformable, pairwise medical image registration. Traditional registration methods optimize an objective function for each pair of images, which can be time-consuming for large datasets or rich deformation models. In contrast to this approach, and building on recent learning-based methods, we formulate registration as a function that maps an input image pair to a deformation field that aligns these images. We parameterize the function via a convolutional neural network (CNN), and optimize the parameters of the neural network on a set of images. Given a new pair of scans, VoxelMorph rapidly computes a deformation field by directly evaluating the function. In this work, we explore two different training strategies. In the first (unsupervised) setting, we train the model to maximize standard image matching objective functions that are based on the image intensities. In the second setting, we leverage auxiliary segmentations available in the training data. We demonstrate that the unsupervised model's accuracy is comparable to state-of-the-art methods, while operating orders of magnitude faster. We also show that VoxelMorph trained with auxiliary data improves registration accuracy at test time, and evaluate the effect of training set size on registration. Our method promises to speed up medical image analysis and processing pipelines, while facilitating novel directions in learning-based registration and its applications. Our code is freely available at voxelmorph.csail.mit.edu.

Guha Balakrishnan, Amy Zhao, Adrian V. Dalca et al.

We address the computational problem of novel human pose synthesis. Given an image of a person and a desired pose, we produce a depiction of that person in that pose, retaining the appearance of both the person and background. We present a modular generative neural network that synthesizes unseen poses using training pairs of images and poses taken from human action videos. Our network separates a scene into different body part and background layers, moves body parts to new locations and refines their appearances, and composites the new foreground with a hole-filled background. These subtasks, implemented with separate modules, are trained jointly using only a single target image as a supervised label. We use an adversarial discriminator to force our network to synthesize realistic details conditioned on pose. We demonstrate image synthesis results on three action classes: golf, yoga/workouts and tennis, and show that our method produces accurate results within action classes as well as across action classes. Given a sequence of desired poses, we also produce coherent videos of actions.

Ronnachai Jaroensri, Amy Zhao, Guha Balakrishnan et al.

For many movement disorders, such as Parkinson's disease and ataxia, disease progression is visually assessed by a clinician using a numerical disease rating scale. These tests are subjective, time-consuming, and must be administered by a professional. This can be problematic where specialists are not available, or when a patient is not consistently evaluated by the same clinician. We present an automated method for quantifying the severity of motion impairment in patients with ataxia, using only video recordings. We consider videos of the finger-to-nose test, a common movement task used as part of the assessment of ataxia progression during the course of routine clinical checkups. Our method uses neural network-based pose estimation and optical flow techniques to track the motion of the patient's hand in a video recording. We extract features that describe qualities of the motion such as speed and variation in performance. Using labels provided by an expert clinician, we train a supervised learning model that predicts severity according to the Brief Ataxia Rating Scale (BARS). The performance of our system is comparable to that of a group of ataxia specialists in terms of mean error and correlation, and our system's predictions were consistently within the range of inter-rater variability. This work demonstrates the feasibility of using computer vision and machine learning to produce consistent and clinically useful measures of motor impairment.