Rahul Sukthankar

Rahul Sukthankar

http://www.cs.cmu.edu/~rahuls/bio.html
Publication list below is partial. For a complete list, please see: http://www.cs.cmu.edu/~rahuls/pub/.
Authored Publications
Sort By
  • Title
  • Title, descending
  • Year
  • Year, descending
    Preview abstract Vision Transformer (ViT) is emerging as the state-of-the-art architecture for image recognition. While recent studies suggest that ViTs are more robust than their convolutional counterparts, our experiments find that ViTs trained on ImageNet are overly reliant on local textures and fail to make adequate use of shape information. ViTs thus have difficulties generalizing to out-of-distribution, real-world data. To address this deficiency, we present a simple and effective architecture modification to ViT's input layer by adding discrete tokens produced by a vector-quantized encoder. Different from the standard continuous pixel tokens, discrete tokens are invariant under small perturbations and contain less information individually, which promote ViTs to learn global information that is invariant. Experimental results demonstrate that adding discrete representation on four architecture variants strengthens ViT robustness by up to 12% across seven ImageNet robustness benchmarks while maintaining the performance on ImageNet. View details
    Neural Descent for Visual 3D Human Pose and Shape
    Andrei Zanfir
    Mihai Zanfir
    Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)(2021), pp. 14484-14493
    Preview abstract We present a deep neural network methodology to reconstruct the 3d pose and shape of people, given image or video inputs. We rely on a recently introduced, expressive full body statistical 3d human model, GHUM, with facial expression and hand detail and aim to learn to reconstruct the model pose and shape states in a self-supervised regime. Central to our methodology, is a learning to learn approach, referred to as HUman Neural Descent (HUND) that avoids both second-order differentiation when training the model parameters, and expensive state gradient descent in order to accurately minimize a semantic differentiable rendering loss at test time. Instead, we rely on novel recurrent stages to update the pose and shape parameters such that not only losses are minimized effectively but the process is regularized in order to ensure progress. The newly introduced architecture is tested extensively, and achieves state-of-the-art results on datasets like H3.6M and 3DPW, as well as in complex imagery collected in-the-wild. View details
    THUNDR: Transformer-based 3D HUmaN Reconstruction with Markers
    Mihai Zanfir
    Andrei Zanfir
    Proceedings of the IEEE/CVF International Conference on Computer Vision(2021)
    Preview abstract We present THUNDR, a transformer-based deep neural network methodology to reconstruct the 3D pose and shape of people, given monocular RGB images. Key to our methodology is an intermediate 3D marker representation, where we aim to combine the predictive power of model-free output architectures and the regularizing, anthropometrically-preserving properties of a statistical human surface models like GHUM—a recently introduced, expressive full body statistical 3d human model, trained end-to-end. Our novel transformer-based prediction pipeline can focus on image regions relevant to the task, supports self-supervised regimes, and ensures that solutions are consistent with human anthropometry. We show state-of-the-art results on Human3.6M and 3DPW, for both the fully-supervised and the self-supervised models, for the task of inferring 3D human shape, joint positions, and global translation. Moreover, we observe very solid 3d reconstruction performance for difficult human poses collected in the wild. Models will be made available for research. View details
    GHUM & GHUML: Generative 3D Human Shape and Articulated Pose Models
    Hongyi Xu
    Andrei Zanfir
    IEEE/CVF Conference on Computer Vision and Pattern Recognition (Oral)(2020), pp. 6184-6193
    Preview abstract We present a statistical, articulated 3D human shape modeling pipeline, within a fully trainable, modular, deep learning framework. Given high-resolution complete 3D body scans of humans, captured in various poses, together with additional closeups of their head and facial expressions, as well as hand articulation, and given initial, artist designed, gender neutral rigged quad-meshes, we train all model parameters including non-linear shape spaces based on variational auto-encoders, pose-space deformation correctives, skeleton joint center predictors, and blend skinning functions, in a single consistent learning loop. The models are simultaneously trained with all the 3d dynamic scan data (over60,000diverse human configurations in our new dataset) in order to capture correlations and en-sure consistency of various components. Models support facial expression analysis, as well as body (with detailed hand) shape and pose estimation. We provide fully train-able generic human models of different resolutions – the moderate-resolution GHUM consisting of 10,168 vertices and the low-resolution GHUML(ite) of 3,194 vertices –, run comparisons between them, analyze the impact of different components and illustrate their reconstruction from image data. The models are available for research. View details
    Preview abstract Can we guess human action from dialogue alone? In this work we investigate the link between spoken words and actions in movies. We note that movie scripts describe actions, as well as contain the speech of characters and hence can be used to learn this correlation with no additional supervision. We train a speech to action classifier on 1k movie scripts downloaded from IMSDb and show that such a classifier performs well for certain classes, and when applied to the speech segments of a large \textit{unlabelled} movie corpus (288k videos, 188M speech segments), provides weak labels for over 800k video clips. By training on these video clips, we demonstrate superior action recognition performance on standard action recognition benchmarks, without using a single labelled action example. View details
    Preview abstract Monocular 3D human pose and shape estimation is challenging due to the many degrees of freedom of the human body and thedifficulty to acquire training data for large-scale supervised learning incomplex visual scenes. In this paper we present practical semi-supervisedand self-supervised models that support training and good generalizationin real-world images and video. Our formulation is based on kinematiclatent normalizing flow representations and dynamics, as well as differ-entiable, semantic body part alignment loss functions that support self-supervised learning. In extensive experiments using 3D motion capturedatasets like CMU, Human3.6M, 3DPW, or AMASS, as well as imagerepositories like COCO, we show that the proposed methods outperformthe state of the art, supporting the practical construction of an accuratefamily of models based on large-scale training with diverse and incom-pletely labeled image and video data. View details
    Preview abstract This paper focuses on multi-person action forecasting in videos. More precisely, given a history of H previous frames, the goal is to detect actors and to predict their future actions for the next T frames. Our approach jointly models temporal and spatial interactions among different actors by constructing a recurrent graph, using actor proposals obtained with Faster R-CNN as nodes. Our method learns to select a subset of discriminative relations without requiring explicit supervision, thus enabling us to tackle challenging visual data. We refer to our model as Discriminative Relational Recurrent Network (DRRN). Evaluation of action prediction on AVA demonstrates the effectiveness of our proposed method compared to simpler baselines. Furthermore, we significantly improve performance on the task of early action classification on J-HMDB, from the previous SOTA of 48% to 60%. View details
    Preview abstract This paper introduces a video dataset of spatio-temporally localized Atomic Visual Actions (AVA). The AVA dataset densely annotates 80 atomic visual actions in 430 15-minute video clips, where actions are localized in space and time, resulting in 1.58M action labels with multiple labels per person occurring frequently. The key characteristics of our dataset are: (1) the definition of atomic visual actions, rather than composite actions; (2) precise spatio-temporal annotations with possibly multiple annotations for each person; (3) exhaustive annotation of these atomic actions over 15-minute video clips; (4) people temporally linked across consecutive segments; and (5) using movies to gather a varied set of action representations. This departs from existing datasets for spatio-temporal action recognition, which typically provide sparse annotations for composite actions in short video clips. We will release the dataset publicly. AVA, with its realistic scene and action complexity, exposes the intrinsic difficulty of action recognition. To benchmark this, we present a novel approach for action localization that builds upon the current state-of-the-art methods, and demonstrates better performance on JHMDB and UCF101-24 categories. While setting a new state of the art on existing datasets, the overall results on AVA are low at 15.6% mAP, underscoring the need for developing new approaches for video understanding. View details
    Preview abstract We propose TAL-Net, an improved approach to temporal action localization in video that is inspired by the Faster R-CNN object detection framework. TAL-Net addresses three key shortcomings of existing approaches: (1) we improve receptive field alignment using a multi-scale architecture that can accommodate extreme variation in action durations; (2) we better exploit the temporal context of actions for both proposal generation and action classification by appropriately extending receptive fields; and (3) we explicitly consider multi-stream feature fusion and demonstrate that fusing motion late is important. We achieve state-of-the-art performance for both action proposal and localiza- tion on THUMOS’14 detection benchmark and competitive performance on ActivityNet challenge. View details
    Preview abstract Current state-of-the-art approaches for spatio-temporal action localization rely on detections at the frame level and model temporal context with 3D ConvNets. Here, we go one step further and model spatio-temporal relations to capture the interactions between human actors, relevant objects and scene elements essential to differentiate similar human actions. Our approach is weakly supervised and mines the relevant elements automatically with an actor-centric relational network (ACRN). ACRN computes and accumulates pair-wise relation information from actor and global scene features, and generates relation features for action classification. It is implemented as neural networks and can be trained jointly with an existing action detection system. We show that ACRN outperforms alternative approaches which capture relation information, and that the proposed framework improves upon the state-of-the-art performance on JHMDB and AVA. A visualization of the learned relation features confirms that our approach is able to attend to the relevant relations for each action. View details