The 3D human pose is vital for modern computer vision and computer graphics, and its prediction has drawn attention in recent years. 3D human pose prediction aims at forecasting a human's future motion from the previous sequence. Ignoring that the arbitrariness of human motion sequences has a firm origin in transition in both temporal and spatial axes limits the performance of state-of-the-art methods, leading them to struggle with making precise predictions on complex cases, e.g., arbitrarily posing or greeting. To alleviate this problem, a network called HaarMoDic is proposed in this paper, which utilizes the 2D Haar transform to project joints to higher resolution coordinates where the network can access spatial and temporal information simultaneously. An ablation study proves that the significant contributing module within the HaarModic Network is the Multi-Resolution Haar (MR-Haar) block. Instead of mining in one of two axes or extracting separately, the MR-Haar block projects whole motion sequences to a mixed-up coordinate in higher resolution with 2D Haar Transform, allowing the network to give scope to information from both axes in different resolutions. With the MR-Haar block, the HaarMoDic network can make predictions referring to a broader range of information. Experimental results demonstrate that HaarMoDic surpasses state-of-the-art methods in every testing interval on the Human3.6M dataset in the Mean Per Joint Position Error (MPJPE) metric.
Human affordance learning investigates contextually relevant novel pose prediction such that the estimated pose represents a valid human action within the scene. While the task is fundamental to machine perception and automated interactive navigation agents, the exponentially large number of probable pose and action variations make the problem challenging and non-trivial. However, the existing datasets and methods for human affordance prediction in 2D scenes are significantly limited in the literature. In this paper, we propose a novel cross-attention mechanism to encode the scene context for affordance prediction by mutually attending spatial feature maps from two different modalities. The proposed method is disentangled among individual subtasks to efficiently reduce the problem complexity. First, we sample a probable location for a person within the scene using a variational autoencoder (VAE) conditioned on the global scene context encoding. Next, we predict a potential pose template from a set of existing human pose candidates using a classifier on the local context encoding around the predicted location. In the subsequent steps, we use two VAEs to sample the scale and deformation parameters for the predicted pose template by conditioning on the local context and template class. Our experiments show significant improvements over the previous baseline of human affordance injection into complex 2D scenes.
The ability of intelligent systems to predict human behaviors is crucial, particularly in fields such as autonomous vehicle navigation and social robotics. However, the complexity of human motion have prevented the development of a standardized dataset for human motion prediction, thereby hindering the establishment of pre-trained models. In this paper, we address these limitations by integrating multiple datasets, encompassing both trajectory and 3D pose keypoints, to propose a pre-trained model for human motion prediction. We merge seven distinct datasets across varying modalities and standardize their formats. To facilitate multimodal pre-training, we introduce Multi-Transmotion, an innovative transformer-based model designed for cross-modality pre-training. Additionally, we present a novel masking strategy to capture rich representations. Our methodology demonstrates competitive performance across various datasets on several downstream tasks, including trajectory prediction in the NBA and JTA datasets, as well as pose prediction in the AMASS and 3DPW datasets. The code is publicly available: https://github.com/vita-epfl/multi-transmotion
Predicting a ligand's bound pose to a target protein is a key component of early-stage computational drug discovery. Recent developments in machine learning methods have focused on improving pose quality at the cost of model runtime. For high-throughput virtual screening applications, this exposes a capability gap that can be filled by moderately accurate but fast pose prediction. To this end, we developed QuickBind, a light-weight pose prediction algorithm. We assess QuickBind on widely used benchmarks and find that it provides an attractive trade-off between model accuracy and runtime. To facilitate virtual screening applications, we augment QuickBind with a binding affinity module and demonstrate its capabilities for multiple clinically-relevant drug targets. Finally, we investigate the mechanistic basis by which QuickBind makes predictions and find that it has learned key physicochemical properties of molecular docking, providing new insights into how machine learning models generate protein-ligand poses. By virtue of its simplicity, QuickBind can serve as both an effective virtual screening tool and a minimal test bed for exploring new model architectures and innovations. Model code and weights are available at https://github.com/aqlaboratory/QuickBind .
Predicting pedestrian behavior is challenging yet crucial for applications such as autonomous driving and smart city. Recent deep learning models have achieved remarkable performance in making accurate predictions, but they fail to provide explanations of their inner workings. One reason for this problem is the multi-modal inputs. To bridge this gap, we present Sparse Prototype Network (SPN), an explainable method designed to simultaneously predict a pedestrian's future action, trajectory, and pose. SPN leverages an intermediate prototype bottleneck layer to provide sample-based explanations for its predictions. The prototypes are modality-independent, meaning that they can correspond to any modality from the input. Therefore, SPN can extend to arbitrary combinations of modalities. Regularized by mono-semanticity and clustering constraints, the prototypes learn consistent and human-understandable features and achieve state-of-the-art performance on action, trajectory and pose prediction on TITAN and PIE. Finally, we propose a metric named Top-K Mono-semanticity Scale to quantitatively evaluate the explainability. Qualitative results show the positive correlation between sparsity and explainability. Code available at https://github.com/Equinoxxxxx/SPN.
The field of protein-ligand pose prediction has seen significant advances in recent years, with machine learning-based methods now being commonly used in lieu of classical docking methods or even to predict all-atom protein-ligand complex structures. Most contemporary studies focus on the accuracy and physical plausibility of ligand placement to determine pose quality, often neglecting a direct assessment of the interactions observed with the protein. In this work, we demonstrate that ignoring protein-ligand interaction fingerprints can lead to overestimation of model performance, most notably in recent protein-ligand cofolding models which often fail to recapitulate key interactions.
Accurately localizing two-dimensional (2D) ultrasound (US) fetal brain images in the 3D brain, using minimal computational resources, is an important task for automated US analysis of fetal growth and development. We propose an uncertainty-aware deep learning model for automated 3D plane localization in 2D fetal brain images. Specifically, a multi-head network is trained to jointly regress 3D plane pose from 2D images in terms of different geometric transformations. The model explicitly learns to predict uncertainty to allocate higher weight to inputs with low variances across different transformations to improve performance. Our proposed method, QAERTS, demonstrates superior pose estimation accuracy than the state-of-the-art and most of the uncertainty-based approaches, leading to 9% improvement on plane angle (PA) for localization accuracy, and 8% on normalized cross-correlation (NCC) for sampled image quality. QAERTS also demonstrates efficiency, containing 5$\times$ fewer parameters than ensemble-based approach, making it advantageous in resource-constrained settings. In addition, QAERTS proves to be more robust to noise effects observed in freehand US scanning by leveraging rotational discontinuities and explicit output uncertainties.
In recent years, machine learning (ML) methods have emerged as promising alternatives for molecular docking, offering the potential for high accuracy without incurring prohibitive computational costs. However, recent studies have indicated that these ML models may overfit to quantitative metrics while neglecting the physical constraints inherent in the problem. In this work, we present Uni-Mol Docking V2, which demonstrates a remarkable improvement in performance, accurately predicting the binding poses of 77+% of ligands in the PoseBusters benchmark with an RMSD value of less than 2.0 {\AA}, and 75+% passing all quality checks. This represents a significant increase from the 62% achieved by the previous Uni-Mol Docking model. Notably, our Uni-Mol Docking approach generates chemically accurate predictions, circumventing issues such as chirality inversions and steric clashes that have plagued previous ML models. Furthermore, we observe enhanced performance in terms of high-quality predictions (RMSD values of less than 1.0 {\AA} and 1.5 {\AA}) and physical soundness when Uni-Mol Docking is combined with more physics-based methods like Uni-Dock. Our results represent a significant advancement in the application of artificial intelligence for scientific research, adopting a holistic approach to ligand docking that is well-suited for industrial applications in virtual screening and drug design. The code, data and service for Uni-Mol Docking are publicly available for use and further development in https://github.com/dptech-corp/Uni-Mol.
We present a method to build animatable dog avatars from monocular videos. This is challenging as animals display a range of (unpredictable) non-rigid movements and have a variety of appearance details (e.g., fur, spots, tails). We develop an approach that links the video frames via a 4D solution that jointly solves for animal's pose variation, and its appearance (in a canonical pose). To this end, we significantly improve the quality of template-based shape fitting by endowing the SMAL parametric model with Continuous Surface Embeddings, which brings image-to-mesh reprojection constaints that are denser, and thus stronger, than the previously used sparse semantic keypoint correspondences. To model appearance, we propose an implicit duplex-mesh texture that is defined in the canonical pose, but can be deformed using SMAL pose coefficients and later rendered to enforce a photometric compatibility with the input video frames. On the challenging CoP3D and APTv2 datasets, we demonstrate superior results (both in terms of pose estimates and predicted appearance) to existing template-free (RAC) and template-based approaches (BARC, BITE).
Protein (receptor)--ligand interaction prediction is a critical component in computer-aided drug design, significantly influencing molecular docking and virtual screening processes. Despite the development of numerous scoring functions in recent years, particularly those employing machine learning, accurately and efficiently predicting binding affinities for protein--ligand complexes remains a formidable challenge. Most contemporary methods are tailored for specific tasks, such as binding affinity prediction, binding pose prediction, or virtual screening, often failing to encompass all aspects. In this study, we put forward DeepRLI, a novel protein--ligand interaction prediction architecture. It encodes each protein--ligand complex into a fully connected graph, retaining the integrity of the topological and spatial structure, and leverages the improved graph transformer layers with cosine envelope as the central module of the neural network, thus exhibiting superior scoring power. In order to equip the model to generalize to conformations beyond the confines of crystal structures and to adapt to molecular docking and virtual screening tasks, we propose a multi-objective strategy, that is, the model outputs three scores for scoring and ranking, docking, and screening, and the training process optimizes these three objectives simultaneously. For the latter two objectives, we augment the dataset through a docking procedure, incorporate suitable physics-informed blocks and employ an effective contrastive learning approach. Eventually, our model manifests a balanced performance across scoring, ranking, docking, and screening, thereby demonstrating its ability to handle a range of tasks. Overall, this research contributes a multi-objective framework for universal protein--ligand interaction prediction, augmenting the landscape of structure-based drug design.
Deep learning promises to dramatically improve scoring functions for molecular docking, leading to substantial advances in binding pose prediction and virtual screening. To train scoring functions-and to perform molecular docking-one must generate a set of candidate ligand binding poses. Unfortunately, the sampling protocols currently used to generate candidate poses frequently fail to produce any poses close to the correct, experimentally determined pose, unless information about the correct pose is provided. This limits the accuracy of learned scoring functions and molecular docking. Here, we describe two improved protocols for pose sampling: GLOW (auGmented sampLing with sOftened vdW potential) and a novel technique named IVES (IteratiVe Ensemble Sampling). Our benchmarking results demonstrate the effectiveness of our methods in improving the likelihood of sampling accurate poses, especially for binding pockets whose shape changes substantially when different ligands bind. This improvement is observed across both experimentally determined and AlphaFold-generated protein structures. Additionally, we present datasets of candidate ligand poses generated using our methods for each of around 5,000 protein-ligand cross-docking pairs, for training and testing scoring functions. To benefit the research community, we provide these cross-docking datasets and an open-source Python implementation of GLOW and IVES at https://github.com/drorlab/GLOW_IVES .
Molecular docking, a key technique in structure-based drug design, plays pivotal roles in protein-ligand interaction modeling, hit identification and optimization, in which accurate prediction of protein-ligand binding mode is essential. Conventional docking approaches perform well in redocking tasks with known protein binding pocket conformation in the complex state. However, in real-world docking scenario without knowing the protein binding conformation for a new ligand, accurately modeling the binding complex structure remains challenging as flexible docking is computationally expensive and inaccurate. Typical deep learning-based docking methods do not explicitly consider protein side chain conformations and fail to ensure the physical plausibility and detailed atomic interactions. In this study, we present DiffBindFR, a full-atom diffusion-based flexible docking model that operates over the product space of ligand overall movements and flexibility and pocket side chain torsion changes. We show that DiffBindFR has higher accuracy in producing native-like binding structures with physically plausible and detailed interactions than available docking methods. Furthermore, in the Apo and AlphaFold2 modeled structures, DiffBindFR demonstrates superior advantages in accurate ligand binding pose and protein binding conformation prediction, making it suitable for Apo and AlphaFold2 structure-based drug design. DiffBindFR provides a powerful flexible docking tool for modeling accurate protein-ligand binding structures.
As the size of accessible compound libraries expands to over 10 billion, the need for more efficient structure-based virtual screening methods is emerging. Different pre-screening methods have been developed for rapid screening, but there is still a lack of structure-based methods applicable to various proteins that perform protein-ligand binding conformation prediction and scoring in an extremely short time. Here, we describe for the first time a deep-learning framework for structure-based pharmacophore modeling to address this challenge. We frame pharmacophore modeling as an instance segmentation problem to determine each protein hotspot and the location of corresponding pharmacophores, and protein-ligand binding pose prediction as a graph-matching problem. PharmacoNet is significantly faster than state-of-the-art structure-based approaches, yet reasonably accurate with a simple scoring function. Furthermore, we show the promising result that PharmacoNet effectively retains hit candidates even under the high pre-screening filtration rates. Overall, our study uncovers the hitherto untapped potential of a pharmacophore modeling approach in deep learning-based drug discovery.
Estimating the head pose of a person is a crucial problem for numerous applications that is yet mainly addressed as a subtask of frontal pose prediction. We present a novel method for unconstrained end-to-end head pose estimation to tackle the challenging task of full range of orientation head pose prediction. We address the issue of ambiguous rotation labels by introducing the rotation matrix formalism for our ground truth data and propose a continuous 6D rotation matrix representation for efficient and robust direct regression. This allows to efficiently learn full rotation appearance and to overcome the limitations of the current state-of-the-art. Together with new accumulated training data that provides full head pose rotation data and a geodesic loss approach for stable learning, we design an advanced model that is able to predict an extended range of head orientations. An extensive evaluation on public datasets demonstrates that our method significantly outperforms other state-of-the-art methods in an efficient and robust manner, while its advanced prediction range allows the expansion of the application area. We open-source our training and testing code along with our trained models: https://github.com/thohemp/6DRepNet360.
Colorectal cancer is one of the most common cancers in the world. While colonoscopy is an effective screening technique, navigating an endoscope through the colon to detect polyps is challenging. A 3D map of the observed surfaces could enhance the identification of unscreened colon tissue and serve as a training platform. However, reconstructing the colon from video footage remains difficult. Learning-based approaches hold promise as robust alternatives, but necessitate extensive datasets. Establishing a benchmark dataset, the 2022 EndoVis sub-challenge SimCol3D aimed to facilitate data-driven depth and pose prediction during colonoscopy. The challenge was hosted as part of MICCAI 2022 in Singapore. Six teams from around the world and representatives from academia and industry participated in the three sub-challenges: synthetic depth prediction, synthetic pose prediction, and real pose prediction. This paper describes the challenge, the submitted methods, and their results. We show that depth prediction from synthetic colonoscopy images is robustly solvable, while pose estimation remains an open research question.
3D perceptual representations are well suited for robot manipulation as they easily encode occlusions and simplify spatial reasoning. Many manipulation tasks require high spatial precision in end-effector pose prediction, which typically demands high-resolution 3D feature grids that are computationally expensive to process. As a result, most manipulation policies operate directly in 2D, foregoing 3D inductive biases. In this paper, we introduce Act3D, a manipulation policy transformer that represents the robot's workspace using a 3D feature field with adaptive resolutions dependent on the task at hand. The model lifts 2D pre-trained features to 3D using sensed depth, and attends to them to compute features for sampled 3D points. It samples 3D point grids in a coarse to fine manner, featurizes them using relative-position attention, and selects where to focus the next round of point sampling. In this way, it efficiently computes 3D action maps of high spatial resolution. Act3D sets a new state-of-the-art in RL-Bench, an established manipulation benchmark, where it achieves 10% absolute improvement over the previous SOTA 2D multi-view policy on 74 RLBench tasks and 22% absolute improvement with 3x less compute over the previous SOTA 3D policy. We quantify the importance of relative spatial attention, large-scale vision-language pre-trained 2D backbones, and weight tying across coarse-to-fine attentions in ablative experiments. Code and videos are available on our project website: https://act3d.github.io/.
Human perception of surroundings is often guided by the various poses present within the environment. Many computer vision tasks, such as human action recognition and robot imitation learning, rely on pose-based entities like human skeletons or robotic arms. However, conventional Vision Transformer (ViT) models uniformly process all patches, neglecting valuable pose priors in input videos. We argue that incorporating poses into RGB data is advantageous for learning fine-grained and viewpoint-agnostic representations. Consequently, we introduce two strategies for learning pose-aware representations in ViTs. The first method, called Pose-aware Attention Block (PAAB), is a plug-and-play ViT block that performs localized attention on pose regions within videos. The second method, dubbed Pose-Aware Auxiliary Task (PAAT), presents an auxiliary pose prediction task optimized jointly with the primary ViT task. Although their functionalities differ, both methods succeed in learning pose-aware representations, enhancing performance in multiple diverse downstream tasks. Our experiments, conducted across seven datasets, reveal the efficacy of both pose-aware methods on three video analysis tasks, with PAAT holding a slight edge over PAAB. Both PAAT and PAAB surpass their respective backbone Transformers by up to 9.8% in real-world action recognition and 21.8% in multi-view robotic video alignment. Code is available at https://github.com/dominickrei/PoseAwareVT.
We introduce the Aria Digital Twin (ADT) - an egocentric dataset captured using Aria glasses with extensive object, environment, and human level ground truth. This ADT release contains 200 sequences of real-world activities conducted by Aria wearers in two real indoor scenes with 398 object instances (324 stationary and 74 dynamic). Each sequence consists of: a) raw data of two monochrome camera streams, one RGB camera stream, two IMU streams; b) complete sensor calibration; c) ground truth data including continuous 6-degree-of-freedom (6DoF) poses of the Aria devices, object 6DoF poses, 3D eye gaze vectors, 3D human poses, 2D image segmentations, image depth maps; and d) photo-realistic synthetic renderings. To the best of our knowledge, there is no existing egocentric dataset with a level of accuracy, photo-realism and comprehensiveness comparable to ADT. By contributing ADT to the research community, our mission is to set a new standard for evaluation in the egocentric machine perception domain, which includes very challenging research problems such as 3D object detection and tracking, scene reconstruction and understanding, sim-to-real learning, human pose prediction - while also inspiring new machine perception tasks for augmented reality (AR) applications. To kick start exploration of the ADT research use cases, we evaluated several existing state-of-the-art methods for object detection, segmentation and image translation tasks that demonstrate the usefulness of ADT as a benchmarking dataset.
We address the task of estimating 6D camera poses from sparse-view image sets (2-8 images). This task is a vital pre-processing stage for nearly all contemporary (neural) reconstruction algorithms but remains challenging given sparse views, especially for objects with visual symmetries and texture-less surfaces. We build on the recent RelPose framework which learns a network that infers distributions over relative rotations over image pairs. We extend this approach in two key ways; first, we use attentional transformer layers to process multiple images jointly, since additional views of an object may resolve ambiguous symmetries in any given image pair (such as the handle of a mug that becomes visible in a third view). Second, we augment this network to also report camera translations by defining an appropriate coordinate system that decouples the ambiguity in rotation estimation from translation prediction. Our final system results in large improvements in 6D pose prediction over prior art on both seen and unseen object categories and also enables pose estimation and 3D reconstruction for in-the-wild objects.
3D interacting hand pose estimation from a single RGB image is a challenging task, due to serious self-occlusion and inter-occlusion towards hands, confusing similar appearance patterns between 2 hands, ill-posed joint position mapping from 2D to 3D, etc.. To address these, we propose to extend A2J-the state-of-the-art depth-based 3D single hand pose estimation method-to RGB domain under interacting hand condition. Our key idea is to equip A2J with strong local-global aware ability to well capture interacting hands' local fine details and global articulated clues among joints jointly. To this end, A2J is evolved under Transformer's non-local encoding-decoding framework to build A2J-Transformer. It holds 3 main advantages over A2J. First, self-attention across local anchor points is built to make them global spatial context aware to better capture joints' articulation clues for resisting occlusion. Secondly, each anchor point is regarded as learnable query with adaptive feature learning for facilitating pattern fitting capacity, instead of having the same local representation with the others. Last but not least, anchor point locates in 3D space instead of 2D as in A2J, to leverage 3D pose prediction. Experiments on challenging InterHand 2.6M demonstrate that, A2J-Transformer can achieve state-of-the-art model-free performance (3.38mm MPJPE advancement in 2-hand case) and can also be applied to depth domain with strong generalization.