We consider the problem of synthesizing multi-action human motion sequences of arbitrary lengths. Existing approaches have mastered motion sequence generation in single action scenarios, but fail to generalize to multi-action and arbitrary-length sequences. We fill this gap by proposing a novel efficient approach that leverages expressiveness of Recurrent Transformers and generative richness of conditional Variational Autoencoders. The proposed iterative approach is able to generate smooth and realistic human motion sequences with an arbitrary number of actions and frames while doing so in linear space and time. We train and evaluate the proposed approach on PROX and Charades datasets, where we augment PROX with ground-truth action labels and Charades with human mesh annotations. Experimental evaluation shows significant improvements in FID score and semantic consistency metrics compared to the state-of-the-art.

We consider the task of 3D pose estimation and tracking of multiple people seen in an arbitrary number of camera feeds. We propose TesseTrack1, a novel top-down approach that simultaneously reasons about multiple individuals’ 3D body joint reconstructions and associations in space and time in a single end-to-end learnable framework. At the core of our approach is a novel spatio-temporal formulation that operates in a common voxelized feature space aggregated from single- or multiple camera views. After a person detection step, a 4D CNN produces short-term person-specific representations which are then linked across time by a differentiable matcher. The linked descriptions are then merged and deconvolved into 3D poses. This joint spatio-temporal formulation contrasts with previous piecewise strategies that treat 2D pose estimation, 2D-to-3D lifting, and 3D pose tracking as independent sub-problems that are error-prone when solved in isolation. Furthermore, unlike previous methods, TesseTrack is robust to changes in the number of camera views and achieves very good results even if a single view is available at inference time. Quantitative evaluation of 3D pose reconstruction accuracy on standard benchmarks shows significant improvements over the state of the art. Evaluation of multi-person articulated 3D pose tracking in our novel evaluation framework demonstrates the superiority of TesseTrack over strong baselines.

Neural architecture search (NAS) approaches aim at automatically finding novel CNN architectures that fit computational constraints while maintaining a good performance on the target platform. We introduce a novel efficient one-shot NAS approach to optimally search for channel numbers, given latency constraints on a specific hardware. We first show that we can use a black-box approach to estimate a realistic latency model for a specific inference platform, without the need for low-level access to the inference computation. Then, we design a pairwise MRF to score any channel configuration and use dynamic programming to efficiently decode the best performing configuration, yielding an optimal solution for the network width search. Finally, we propose an adaptive channel configuration sampling scheme to gradually specialize the training phase to the target computational constraints. Experiments on ImageNet classification show that our approach can find networks fitting the resource constraints on different target platforms while improving accuracy over the state-of-the-art efficient networks.

We evaluate the robustness of different edge configurations as shown in Figure 2. This is motivated by the fact that edge configuration a) is prone to errors. If a single edge is not estimated correctly, the entire pose breaks. Similar to [4] we introduce skip connections to the standard model (Figure 2 b) Bypass model). Figure 2 c) illustrates a different idea to connect joints which we refer to as Range of Motion (ROM) model since pairs of joints are connected if both lie within the same ROM of a third joint. Further we train an edge configuration as proposed in [2] which we refer to as Extended model. For completeness, we introduce a nearly-fully-connected (NFC) model (Figure 2 e)) which connects most nearby

Existing systems for video-based pose estimation and tracking struggle to perform well on realistic videos with multiple people and often fail to output body-pose trajectories consistent over time. To address this shortcoming this paper introduces PoseTrack which is a new large-scale benchmark for video-based human pose estimation and articulated tracking. Our new benchmark encompasses three tasks focusing on i) single-frame multi-person pose estimation, ii) multi-person pose estimation in videos, and iii) multi-person articulated tracking. To establish the benchmark, we collect, annotate and release a new dataset that features videos with multiple people labeled with person tracks and articulated pose. A public centralized evaluation server is provided to allow the research community to evaluate on a held-out test set. Furthermore, we conduct an extensive experimental study on recent approaches to articulated pose tracking and provide analysis of the strengths and weaknesses of the state of the art. We envision that the proposed benchmark will stimulate productive research both by providing a large and representative training dataset as well as providing a platform to objectively evaluate and compare the proposed methods. The benchmark is freely accessible at https://posetrack.net/.

Multi-person pose estimation in images and videos is an important yet challenging task with many applications. Despite the large improvements in human pose estimation enabled by the development of convolutional neural networks, there still exist a lot of difficult cases where even the state-of-the-art models fail to correctly localize all body joints. This motivates the need for an additional refinement step that addresses these challenging cases and can be easily applied on top of any existing method. In this work, we introduce a pose refinement network (PoseRefiner) which takes as input both the image and a given pose estimate and learns to directly predict a refined pose by jointly reasoning about the input-output space. In order for the network to learn to refine incorrect body joint predictions, we employ a novel data augmentation scheme for training, where we model "hard" human pose cases. We evaluate our approach on four popular large-scale pose estimation benchmarks such as MPII Single-and Multi-Person Pose Estimation, PoseTrack Pose Estimation, and PoseTrack Pose Tracking, and report systematic improvement over the state of the art.

•Expressive 3D human shape models are proposed.•The models are learned from the largest available dataset of laser scans.•Various template fitting and posture normalization approaches are evaluated.•High quality of the learned shape spaces is empirically demonstrated.•Proposed models and code to data pre-processing and model fitting are released.

In this paper we propose an approach for articulated tracking of multiple people in unconstrained videos. Our starting point is a model that resembles existing architectures for single-frame pose estimation but is substantially faster. We achieve this in two ways: (1) by simplifying and sparsifying the body-part relationship graph and leveraging recent methods for faster inference, and (2) by offloading a substantial share of computation onto a feed-forward convolutional architecture that is able to detect and associate body joints of the same person even in clutter. We use this model to generate proposals for body joint locations and formulate articulated tracking as spatio-temporal grouping of such proposals. This allows to jointly solve the association problem for all people in the scene by propagating evidence from strong detections through time and enforcing constraints that each proposal can be assigned to one person only. We report results on a public MPII Human Pose benchmark and on a new MPII Video Pose dataset of image sequences with multiple people. We demonstrate that our model achieves state-of-the-art results while using only a fraction of time and is able to leverage temporal information to improve state-of-the-art for crowded scenes.

The goal of this paper is to advance the state-of-the-art of articulated pose estimation in scenes with multiple people. To that end we contribute on three fronts. We propose (1) improved body part detectors that generate effective bottom-up proposals for body parts; (2) novel image-conditioned pairwise terms that allow to assemble the proposals into a variable number of consistent body part configurations; and (3) an incremental optimization strategy that explores the search space more efficiently thus leading both to better performance and significant speed-up factors. Evaluation is done on two single-person and two multi-person pose estimation benchmarks. The proposed approach significantly outperforms best known multi-person pose estimation results while demonstrating competitive performance on the task of single person pose estimation. Models and code available at http://pose.mpi-inf.mpg.de

In this thesis we are interested in the problem of articulated people detection and pose estimation being key ingredients towards understanding visual scenes containing people. Although extensive efforts are being made to address these problems, we identify three promising directions that, we believe, didn’t get sufficient attention recently. First, we investigate how statistical 3D human shape models from computer graphics can be leveraged to ease training data generation. We propose a range of automatic data generation techniques that allow to directly represent relevant variations in the training data. Sampling from both the underlying human shape distribution and a large dataset of human poses allows to generate novel samples with controllable shape and pose variations that are relevant for the task at hand. Furthermore, we improve the state-of-the-art 3D human shape model itself by rebuilding it from a large commercially available dataset of 3D bodies. Second, we develop expressive spatial and strong appearance models for 2D singleand multi-person pose estimation. We propose an expressive single person model that incorporates higher order part dependencies while remaining efficient. We augment this model with various types of strong appearance representations aiming to substantially improve the body part hypotheses. Finally, we propose an expressive model for joint pose estimation of multiple people. To that end, we develop strong deep learning based body part detectors and an expressive fully connected spatial model. The proposed approach treats multi-person pose estimation as a joint partitioning and labeling problem of a set of body part hypotheses: it infers the number of persons in a scene, identifies occluded body parts and disambiguates body parts between people in close proximity of each other. Third, we perform thorough evaluation and performance analysis of leading human pose estimation and activity recognition methods. To that end we introduce a novel benchmark that makes a significant advance in terms of diversity and difficulty, compared to the previous datasets, and includes over 40, 000 annotated body poses and over 1.5M frames. Furthermore, we provide a rich set of labels which are used to perform a detailed analysis of competing approaches gaining insights into successes and failures of these methods. In summary, this thesis presents a novel approach to articulated people detection and pose estimation. Thorough experimental evaluation on standard benchmarks demonstrates significant improvements due to the proposed data augmentation techniques and novel body models, while detailed performance analysis of competing approaches on our newly introduced large-scale benchmark allows to identify the most promising directions of improvement.

Marker-less motion capture has seen great progress, but most state-of-the-art approaches fail to reliably track articulated human body motion with a very low number of cameras, let alone when applied in outdoor scenes with general background. In this paper, we propose a method for accurate marker-less capture of articulated skeleton motion of several subjects in general scenes, indoors and outdoors, even from input filmed with as few as two cameras. The new algorithm combines the strengths of a discriminative image-based joint detection method with a model-based generative motion tracking algorithm through an unified pose optimization energy. The discriminative part-based pose detection method is implemented using Convolutional Networks (ConvNet) and estimates unary potentials for each joint of a kinematic skeleton model. These unary potentials serve as the basis of a probabilistic extraction of pose constraints for tracking by using weighted sampling from a pose posterior that is guided by the model. In the final energy, we combine these constraints with an appearance-based model-to-image similarity term. Poses can be computed very efficiently using iterative local optimization, since joint detection with a trained ConvNet is fast, and since our formulation yields a combined pose estimation energy with analytic derivatives. In combination, this enables to track full articulated joint angles at state-of-the-art accuracy and temporal stability with a very low number of cameras. Our method is efficient and lends itself to implementation on parallel computing hardware, such as GPUs. We test our method extensively and show its advantages over related work on many indoor and outdoor data sets captured by ourselves, as well as data sets made available to the community by other research labs. The availability of good evaluation data sets is paramount for scientific progress, and many existing test data sets focus on controlled indoor settings, do not feature much variety in the scenes, and often lack a large corpus of data with ground truth annotation. We therefore further contribute with a new extensive test data set called MPI-MARCOnI for indoor and outdoor marker-less motion capture that features 12 scenes of varying complexity and varying camera count, and that features ground truth reference data from different modalities, ranging from manual joint annotations to marker-based motion capture results. Our new method is tested on these data, and the data set will be made available to the community.

This paper considers the task of articulated human pose estimation of multiple people in real world images. We propose an approach that jointly solves the tasks of detection and pose estimation: it infers the number of persons in a scene, identifies occluded body parts, and disambiguates body parts between people in close proximity of each other. This joint formulation is in contrast to previous strategies, that address the problem by first detecting people and subsequently estimating their body pose. We propose a partitioning and labeling formulation of a set of body-part hypotheses generated with CNN-based part detectors. Our formulation, an instance of an integer linear program, implicitly performs non-maximum suppression on the set of part candidates and groups them to form configurations of body parts respecting geometric and appearance constraints. Experiments on four different datasets demonstrate state-of-the-art results for both single person and multi person pose estimation.

We present a novel method for accurate marker-less capture of articulated skeleton motion of several subjects in general scenes, indoors and outdoors, even from input filmed with as few as two cameras. Our approach unites a discriminative image-based joint detection method with a model-based generative motion tracking algorithm through a combined pose optimization energy. The discriminative part-based pose detection method, implemented using Convolutional Networks (ConvNet), estimates unary potentials for each joint of a kinematic skeleton model. These unary potentials are used to probabilistically extract pose constraints for tracking by using weighted sampling from a pose posterior guided by the model. In the final energy, these constraints are combined with an appearance-based model-to-image similarity term. Poses can be computed very efficiently using iterative local optimization, as ConvNet detection is fast, and our formulation yields a combined pose estimation energy with analytic derivatives. In combination, this enables to track full articulated joint angles at state-of-the-art accuracy and temporal stability with a very low number of cameras.

Human pose estimation has made significant progress during the last years. However current datasets are limited in their coverage of the overall pose estimation challenges. Still these serve as the common sources to evaluate, train and compare different models on. In this paper we introduce a novel benchmark "MPII Human Pose" that makes a significant advance in terms of diversity and difficulty, a contribution that we feel is required for future developments in human body models. This comprehensive dataset was collected using an established taxonomy of over 800 human activities [1]. The collected images cover a wider variety of human activities than previous datasets including various recreational, occupational and householding activities, and capture people from a wider range of viewpoints. We provide a rich set of labels including positions of body joints, full 3D torso and head orientation, occlusion labels for joints and body parts, and activity labels. For each image we provide adjacent video frames to facilitate the use of motion information. Given these rich annotations we perform a detailed analysis of leading human pose estimation approaches and gaining insights for the success and failures of these methods.

Estimating the body shape and posture of a dressed human subject in motion represented as a sequence of (possibly incomplete) 3D meshes is important for virtual change rooms and security. To solve this problem, statistical shape spaces encoding human body shape and posture variations are commonly used to constrain the search space for the shape estimate. In this work, we propose a novel method that uses a posture-invariant shape space to model body shape variation combined with a skeleton-based deformation to model posture variation. Our method can estimate the body shape and posture of both static scans and motion sequences of human body scans with clothing that fits relatively closely to the body. In case of motion sequences, our method takes advantage of motion cues to solve for a single body shape estimate along with a sequence of posture estimates. We apply our approach to both static scans and motion sequences and demonstrate that using our method, higher fitting accuracy is achieved than when using a variant of the popular SCAPE model [2], [18] as statistical model.

In this work we consider the challenging task of articulated human pose estimation in monocular images. Most of current methods in this area [4, 8, 16, 14] are based on the pictorial structures model (PS) and are composed of unary terms modelling body part appearance and pairwise terms between adjacent body parts and/or joints capturing their preferred spatial arrangement. In this work we advance the state of the art in articulated human pose estimation in three ways. First, we argue that modeling part dependencies between non-adjacent body parts is important for effective pose estimation (cf. Fig. 1). We propose a model [10] that incorporates higher order information between body parts by defining a conditional model in which all parts are a-priori connected, but which becomes a tractable PS model once the mid-level features are observed. This allows to effectively model dependencies between non-adjacent parts and retains an exact and efficient inference procedure in a tree-based model. Second, we explore various types of appearance representations with the aim to improve the body part hypotheses [11]. We argue that in order to obtain effective part detectors it is necessary to leverage both the pose specific appearance of body parts and the joint appearance of part constellations. We show that the proposed appearance representations are complementary and a combination of the best performing appearance model paired with a flexible image-conditioned spatial model achieves the best result. Third, we introduce a novel benchmark “MPII Human Pose” [3] that makes a significant advance in terms of diversity and difficulty, a contribution that we feel is required for future developments in human body models. This comprehensive dataset was collected using an established taxonomy of over 800 human activities. The collected images cover a wider variety of human activities than previous datasets including various recreational, occupational, and householding activities. People are captured from a wider range of viewpoints. In addition we provide a rich set of labels including positions of body joints, full 3D torso and head orientation, occlusion labels for joints and body parts, and activity labels. With these annotations we perform a detailed analysis [3, 12] of the leading 2D human pose estimation and activity recognition methods to understand success and failure cases for established models.

Holistic methods based on dense trajectories [29, 30] are currently the de facto standard for recognition of human activities in video. Whether holistic representations will sustain or will be superseded by higher level video encoding in terms of body pose and motion is the subject of an ongoing debate [12]. In this paper we aim to clarify the underlying factors responsible for good performance of holistic and pose-based representations. To that end we build on our recent dataset [2] leveraging the existing taxonomy of human activities. This dataset includes 24,920 video snippets covering 410 human activities in total. Our analysis reveals that holistic and pose-based methods are highly complementary, and their performance varies significantly depending on the activity. We find that holistic methods are mostly affected by the number and speed of trajectories, whereas pose-based methods are mostly influenced by viewpoint of the person. We observe striking performance differences across activities: for certain activities results with pose-based features are more than twice as accurate compared to holistic features, and vice versa. The best performing approach in our comparison is based on the combination of holistic and pose-based approaches, which again underlines their complementarity.

Typical approaches to articulated pose estimation combine spatial modelling of the human body with appearance modelling of body parts. This paper aims to push the state-of-the-art in articulated pose estimation in two ways. First we explore various types of appearance representations aiming to substantially improve the body part hypotheses. And second, we draw on and combine several recently proposed powerful ideas such as more flexible spatial models as well as image-conditioned spatial models. In a series of experiments we draw several important conclusions: (1) we show that the proposed appearance representations are complementary, (2) we demonstrate that even a basic tree-structure spatial human body model achieves state-of-the-art performance when augmented with the proper appearance representation, and (3) we show that the combination of the best performing appearance model with a flexible image-conditioned spatial model achieves the best result, significantly improving over the state of the art, on the ``Leeds Sports Poses'' and ``Parse'' benchmarks.

In this paper we consider the challenging problem of articulated human pose estimation in still images. We observe that despite high variability of the body articulations, human motions and activities often simultaneously constrain the positions of multiple body parts. Modelling such higher order part dependencies seemingly comes at a cost of more expensive inference, which resulted in their limited use in state-of-the-art methods. In this paper we propose a model that incorporates higher order part dependencies while remaining efficient. We achieve this by defining a conditional model in which all body parts are connected a-priori, but which becomes a tractable tree-structured pictorial structures model once the image observations are available. In order to derive a set of conditioning variables we rely on the poselet-based features that have been shown to be effective for people detection but have so far found limited application for articulated human pose estimation. We demonstrate the effectiveness of our approach on three publicly available pose estimation benchmarks improving or being on-par with state of the art in each case.

This paper systematically analyzes the strengths and weaknesses of existing image warping algorithms on the tasks of face recognition. Image warping is used to cope with local and global image variability and in general is an NP-complete problem. Although many approximations have recently been proposed, neither thorough comparison, nor systematic analysis of methods in a common scheme has been done so far. We follow the bottom-up approach and analyze the methods with increasing degree of image structure preserved during optimization. We evaluate the presented warping approaches on four challenging face recognition tasks in highly variable domains. Our findings indicate that preserving maximum dependencies between neighboring pixels by imposing strong geometrical constraints leads to the best recognition results while making optimization efficient.