Searcharxiv⌕ Search

arXiv subjects

Jiajun Wu

Publications and source records attributed to Jiajun Wu.

At least 271 records · Page 15Linked to original sources
arXiv2023

Motion Question Answering via Modular Motion Programs

In order to build artificial intelligence systems that can perceive and reason with human behavior in the real world, we must first design models that conduct complex spatio-temporal reasoning over motion sequences. Moving towards this goal, we propose the HumanMotionQA task to evaluate complex, multi-step reasoning abilities of models on long-form human motion sequences. We generate a dataset of question-answer pairs that require detecting motor cues in small portions of motion sequences, reasoning temporally about when events occur, and querying specific motion attributes. In addition, we propose NSPose, a neuro-symbolic method for this task that uses symbolic reasoning and a modular design to ground motion through learning motion concepts, attribute neural operators, and temporal relations. We demonstrate the suitability of NSPose for the HumanMotionQA task, outperforming all baseline methods.

cs.CV↗
arXiv2023

An Extensible Multimodal Multi-task Object Dataset with Materials

We present EMMa, an Extensible, Multimodal dataset of Amazon product listings that contains rich Material annotations. It contains more than 2.8 million objects, each with image(s), listing text, mass, price, product ratings, and position in Amazon's product-category taxonomy. We also design a comprehensive taxonomy of 182 physical materials (e.g., Plastic $\rightarrow$ Thermoplastic $\rightarrow$ Acrylic). Objects are annotated with one or more materials from this taxonomy. With the numerous attributes available for each object, we develop a Smart Labeling framework to quickly add new binary labels to all objects with very little manual labeling effort, making the dataset extensible. Each object attribute in our dataset can be included in either the model inputs or outputs, leading to combinatorial possibilities in task configurations. For example, we can train a model to predict the object category from the listing text, or the mass and price from the product listing image. EMMa offers a new benchmark for multi-task learning in computer vision and NLP, and allows practitioners to efficiently add new tasks and object attributes at scale.

cs.CV↗
arXiv2023

Putting People in Their Place: Affordance-Aware Human Insertion into Scenes

We study the problem of inferring scene affordances by presenting a method for realistically inserting people into scenes. Given a scene image with a marked region and an image of a person, we insert the person into the scene while respecting the scene affordances. Our model can infer the set of realistic poses given the scene context, re-pose the reference person, and harmonize the composition. We set up the task in a self-supervised fashion by learning to re-pose humans in video clips. We train a large-scale diffusion model on a dataset of 2.4M video clips that produces diverse plausible poses while respecting the scene context. Given the learned human-scene composition, our model can also hallucinate realistic people and scenes when prompted without conditioning and also enables interactive editing. A quantitative evaluation shows that our method synthesizes more realistic human appearance and more natural human-scene interactions than prior work.

cs.CV↗
arXiv2023

A Control-Centric Benchmark for Video Prediction

Video is a promising source of knowledge for embodied agents to learn models of the world's dynamics. Large deep networks have become increasingly effective at modeling complex video data in a self-supervised manner, as evaluated by metrics based on human perceptual similarity or pixel-wise comparison. However, it remains unclear whether current metrics are accurate indicators of performance on downstream tasks. We find empirically that for planning robotic manipulation, existing metrics can be unreliable at predicting execution success. To address this, we propose a benchmark for action-conditioned video prediction in the form of a control benchmark that evaluates a given model for simulated robotic manipulation through sampling-based planning. Our benchmark, Video Prediction for Visual Planning ($VP^2$), includes simulated environments with 11 task categories and 310 task instance definitions, a full planning implementation, and training datasets containing scripted interaction trajectories for each task category. A central design goal of our benchmark is to expose a simple interface -- a single forward prediction call -- so it is straightforward to evaluate almost any action-conditioned video prediction model. We then leverage our benchmark to study the effects of scaling model size, quantity of training data, and model ensembling by analyzing five highly-performant video prediction models, finding that while scale can improve perceptual quality when modeling visually diverse settings, other attributes such as uncertainty awareness can also aid planning performance.

cs.CV↗
arXiv2023

Programmatically Grounded, Compositionally Generalizable Robotic Manipulation

Robots operating in the real world require both rich manipulation skills as well as the ability to semantically reason about when to apply those skills. Towards this goal, recent works have integrated semantic representations from large-scale pretrained vision-language (VL) models into manipulation models, imparting them with more general reasoning capabilities. However, we show that the conventional pretraining-finetuning pipeline for integrating such representations entangles the learning of domain-specific action information and domain-general visual information, leading to less data-efficient training and poor generalization to unseen objects and tasks. To this end, we propose ProgramPort, a modular approach to better leverage pretrained VL models by exploiting the syntactic and semantic structures of language instructions. Our framework uses a semantic parser to recover an executable program, composed of functional modules grounded on vision and action across different modalities. Each functional module is realized as a combination of deterministic computation and learnable neural networks. Program execution produces parameters to general manipulation primitives for a robotic end-effector. The entire modular network can be trained with end-to-end imitation learning objectives. Experiments show that our model successfully disentangles action and perception, translating to improved zero-shot and compositional generalization in a variety of manipulation behaviors. Project webpage at: \url{https://progport.github.io}.

cs.AI↗
arXiv2023

Neurosymbolic Models for Computer Graphics

Procedural models (i.e. symbolic programs that output visual data) are a historically-popular method for representing graphics content: vegetation, buildings, textures, etc. They offer many advantages: interpretable design parameters, stochastic variations, high-quality outputs, compact representation, and more. But they also have some limitations, such as the difficulty of authoring a procedural model from scratch. More recently, AI-based methods, and especially neural networks, have become popular for creating graphic content. These techniques allow users to directly specify desired properties of the artifact they want to create (via examples, constraints, or objectives), while a search, optimization, or learning algorithm takes care of the details. However, this ease of use comes at a cost, as it's often hard to interpret or manipulate these representations. In this state-of-the-art report, we summarize research on neurosymbolic models in computer graphics: methods that combine the strengths of both AI and symbolic programs to represent, generate, and manipulate visual data. We survey recent work applying these techniques to represent 2D shapes, 3D shapes, and materials & textures. Along the way, we situate each prior work in a unified design space for neurosymbolic models, which helps reveal underexplored areas and opportunities for future research.

cs.GR↗
arXiv2023

CIRCLE: Capture In Rich Contextual Environments

Synthesizing 3D human motion in a contextual, ecological environment is important for simulating realistic activities people perform in the real world. However, conventional optics-based motion capture systems are not suited for simultaneously capturing human movements and complex scenes. The lack of rich contextual 3D human motion datasets presents a roadblock to creating high-quality generative human motion models. We propose a novel motion acquisition system in which the actor perceives and operates in a highly contextual virtual world while being motion captured in the real world. Our system enables rapid collection of high-quality human motion in highly diverse scenes, without the concern of occlusion or the need for physical scene construction in the real world. We present CIRCLE, a dataset containing 10 hours of full-body reaching motion from 5 subjects across nine scenes, paired with ego-centric information of the environment represented in various forms, such as RGBD videos. We use this dataset to train a model that generates human motion conditioned on scene information. Leveraging our dataset, the model learns to use ego-centric scene information to achieve nontrivial reaching tasks in the context of complex 3D scenes. To download the data please visit https://stanford-tml.github.io/circle_dataset/.

cs.CV↗
arXiv2023

PyPose: A Library for Robot Learning with Physics-based Optimization

Deep learning has had remarkable success in robotic perception, but its data-centric nature suffers when it comes to generalizing to ever-changing environments. By contrast, physics-based optimization generalizes better, but it does not perform as well in complicated tasks due to the lack of high-level semantic information and reliance on manual parametric tuning. To take advantage of these two complementary worlds, we present PyPose: a robotics-oriented, PyTorch-based library that combines deep perceptual models with physics-based optimization. PyPose's architecture is tidy and well-organized, it has an imperative style interface and is efficient and user-friendly, making it easy to integrate into real-world robotic applications. Besides, it supports parallel computing of any order gradients of Lie groups and Lie algebras and $2^{\text{nd}}$-order optimizers, such as trust region methods. Experiments show that PyPose achieves more than $10\times$ speedup in computation compared to the state-of-the-art libraries. To boost future research, we provide concrete examples for several fields of robot learning, including SLAM, planning, control, and inertial navigation.

cs.RO↗
arXiv2023

NS3D: Neuro-Symbolic Grounding of 3D Objects and Relations

Grounding object properties and relations in 3D scenes is a prerequisite for a wide range of artificial intelligence tasks, such as visually grounded dialogues and embodied manipulation. However, the variability of the 3D domain induces two fundamental challenges: 1) the expense of labeling and 2) the complexity of 3D grounded language. Hence, essential desiderata for models are to be data-efficient, generalize to different data distributions and tasks with unseen semantic forms, as well as ground complex language semantics (e.g., view-point anchoring and multi-object reference). To address these challenges, we propose NS3D, a neuro-symbolic framework for 3D grounding. NS3D translates language into programs with hierarchical structures by leveraging large language-to-code models. Different functional modules in the programs are implemented as neural networks. Notably, NS3D extends prior neuro-symbolic visual reasoning methods by introducing functional modules that effectively reason about high-arity relations (i.e., relations among more than two objects), key in disambiguating objects in complex 3D scenes. Modular and compositional architecture enables NS3D to achieve state-of-the-art results on the ReferIt3D view-dependence task, a 3D referring expression comprehension benchmark. Importantly, NS3D shows significantly improved performance on settings of data-efficiency and generalization, and demonstrate zero-shot transfer to an unseen 3D question-answering task.

cs.CV↗
arXiv2023

Learning Vortex Dynamics for Fluid Inference and Prediction

We propose a novel differentiable vortex particle (DVP) method to infer and predict fluid dynamics from a single video. Lying at its core is a particle-based latent space to encapsulate the hidden, Lagrangian vortical evolution underpinning the observable, Eulerian flow phenomena. Our differentiable vortex particles are coupled with a learnable, vortex-to-velocity dynamics mapping to effectively capture the complex flow features in a physically-constrained, low-dimensional space. This representation facilitates the learning of a fluid simulator tailored to the input video that can deliver robust, long-term future predictions. The value of our method is twofold: first, our learned simulator enables the inference of hidden physics quantities (e.g., velocity field) purely from visual observation; secondly, it also supports future prediction, constructing the input video's sequel along with its future dynamics evolution. We compare our method with a range of existing methods on both synthetic and real-world videos, demonstrating improved reconstruction quality, visual plausibility, and physical integrity.

cs.LG↗
arXiv2023

Differentiable Physics Simulation of Dynamics-Augmented Neural Objects

We present a differentiable pipeline for simulating the motion of objects that represent their geometry as a continuous density field parameterized as a deep network. This includes Neural Radiance Fields (NeRFs), and other related models. From the density field, we estimate the dynamical properties of the object, including its mass, center of mass, and inertia matrix. We then introduce a differentiable contact model based on the density field for computing normal and friction forces resulting from collisions. This allows a robot to autonomously build object models that are visually and \emph{dynamically} accurate from still images and videos of objects in motion. The resulting Dynamics-Augmented Neural Objects (DANOs) are simulated with an existing differentiable simulation engine, Dojo, interacting with other standard simulation objects, such as spheres, planes, and robots specified as URDFs. A robot can use this simulation to optimize grasps and manipulation trajectories of neural objects, or to improve the neural object models through gradient-based real-to-simulation transfer. We demonstrate the pipeline to learn the coefficient of friction of a bar of soap from a real video of the soap sliding on a table. We also learn the coefficient of friction and mass of a Stanford bunny through interactions with a Panda robot arm from synthetic data, and we optimize trajectories in simulation for the Panda arm to push the bunny to a goal location.

cs.RO↗
arXiv2023

Learning Rational Subgoals from Demonstrations and Instructions

We present a framework for learning useful subgoals that support efficient long-term planning to achieve novel goals. At the core of our framework is a collection of rational subgoals (RSGs), which are essentially binary classifiers over the environmental states. RSGs can be learned from weakly-annotated data, in the form of unsegmented demonstration trajectories, paired with abstract task descriptions, which are composed of terms initially unknown to the agent (e.g., collect-wood then craft-boat then go-across-river). Our framework also discovers dependencies between RSGs, e.g., the task collect-wood is a helpful subgoal for the task craft-boat. Given a goal description, the learned subgoals and the derived dependencies facilitate off-the-shelf planning algorithms, such as A* and RRT, by setting helpful subgoals as waypoints to the planner, which significantly improves performance-time efficiency.

cs.AI↗
arXiv2023

Sparse and Local Networks for Hypergraph Reasoning

Reasoning about the relationships between entities from input facts (e.g., whether Ari is a grandparent of Charlie) generally requires explicit consideration of other entities that are not mentioned in the query (e.g., the parents of Charlie). In this paper, we present an approach for learning to solve problems of this kind in large, real-world domains, using sparse and local hypergraph neural networks (SpaLoc). SpaLoc is motivated by two observations from traditional logic-based reasoning: relational inferences usually apply locally (i.e., involve only a small number of individuals), and relations are usually sparse (i.e., only hold for a small percentage of tuples in a domain). We exploit these properties to make learning and inference efficient in very large domains by (1) using a sparse tensor representation for hypergraph neural networks, (2) applying a sparsification loss during training to encourage sparse representations, and (3) subsampling based on a novel information sufficiency-based sampling process during training. SpaLoc achieves state-of-the-art performance on several real-world, large-scale knowledge graph reasoning benchmarks, and is the first framework for applying hypergraph neural networks on real-world knowledge graphs with more than 10k nodes.

cs.LG↗
arXiv2023

DyBit: Dynamic Bit-Precision Numbers for Efficient Quantized Neural Network Inference

To accelerate the inference of deep neural networks (DNNs), quantization with low-bitwidth numbers is actively researched. A prominent challenge is to quantize the DNN models into low-bitwidth numbers without significant accuracy degradation, especially at very low bitwidths (< 8 bits). This work targets an adaptive data representation with variable-length encoding called DyBit. DyBit can dynamically adjust the precision and range of separate bit-field to be adapted to the DNN weights/activations distribution. We also propose a hardware-aware quantization framework with a mixed-precision accelerator to trade-off the inference accuracy and speedup. Experimental results demonstrate that the inference accuracy via DyBit is 1.997% higher than the state-of-the-art at 4-bit quantization, and the proposed framework can achieve up to 8.1x speedup compared with the original model.

cs.LG↗
arXiv2023

VQ3D: Learning a 3D-Aware Generative Model on ImageNet

Recent work has shown the possibility of training generative models of 3D content from 2D image collections on small datasets corresponding to a single object class, such as human faces, animal faces, or cars. However, these models struggle on larger, more complex datasets. To model diverse and unconstrained image collections such as ImageNet, we present VQ3D, which introduces a NeRF-based decoder into a two-stage vector-quantized autoencoder. Our Stage 1 allows for the reconstruction of an input image and the ability to change the camera position around the image, and our Stage 2 allows for the generation of new 3D scenes. VQ3D is capable of generating and reconstructing 3D-aware images from the 1000-class ImageNet dataset of 1.2 million training images. We achieve an ImageNet generation FID score of 16.8, compared to 69.8 for the next best baseline method.

cs.CV↗
arXiv2023

FedLE: Federated Learning Client Selection with Lifespan Extension for Edge IoT Networks

Federated learning (FL) is a distributed and privacy-preserving learning framework for predictive modeling with massive data generated at the edge by Internet of Things (IoT) devices. One major challenge preventing the wide adoption of FL in IoT is the pervasive power supply constraints of IoT devices due to the intensive energy consumption of battery-powered clients for local training and model updates. Low battery levels of clients eventually lead to their early dropouts from edge networks, loss of training data jeopardizing the performance of FL, and their availability to perform other designated tasks. In this paper, we propose FedLE, an energy-efficient client selection framework that enables lifespan extension of edge IoT networks. In FedLE, the clients first run for a minimum epoch to generate their local model update. The models are partially uploaded to the server for calculating similarities between each pair of clients. Clustering is performed against these client pairs to identify those with similar model distributions. In each round, low-powered clients have a lower probability of being selected, delaying the draining of their batteries. Empirical studies show that FedLE outperforms baselines on benchmark datasets and lasts more training rounds than FedAvg with battery power constraints.

cs.LG↗
arXiv2023

$W$-Boson Mass Anomaly from a General $SU(2)_{L}$ Scalar Multiplet

We explain the $W$-boson mass anomaly by introducing an $SU(2)_L$ scalar multiplet with general isospin and hypercharge in the case without its vacuum expectation value. It is shown that the dominant contribution from the scalar multiplet to the $W$-boson mass arises at one-loop level, which can be expressed in terms of the electroweak (EW) oblique parameters $T$ and $S$ at leading order. We firstly rederive the general formulae of $T$ and $S$ induced by a scalar multiplet of EW charges, confirming the results in the literature. We then study several specific examples of great phenomenological interest by applying these general expressions. As a result, it is found that the model with a scalar multiplet in an $SU(2)_L$ real representation with $Y=0$ cannot generate the required $M_W$ correction since it leads to vanishing values of $T$ and $S$. On the other hand, the cases with scalars in a complex representation under $SU(2)_L$ with a general hypercharge can explain the $M_W$ excess observed by CDF-\uppercase\expandafter{\romannumeral2} due to nonzero $T$ and $S$. We further take into account of the strong constraints from the perturbativity and the EW global fit of the precision data, and vary the isospin representation and hypercharge of the additional scalar multiplet, in order to assess the extent of the model to solve the $W$-boson mass anomaly. It turns out that these constraints play important roles in setting limits on the model parameter space. We also briefly describe the collider signatures of the extra scalar multiplet, especially when it contains long-lived heavy highly charged states.

hep-ph↗
arXiv2023

Muon $g-2$ Anomaly from a Massive Spin-2 Particle

We investigate the possibility to interpret the muon $g-2$ anomaly in terms of a massive spin-2 particle, $G$, which can be identified as the first Kaluza-Klein graviton in the generalized Randall-Sundrum model. In particular, we obtain the leading-order contributions to the muon $g-2$ by calculating the relevant one-loop Feynman diagrams induced by $G$. The analytic expression is shown to keep the gauge invariance of the quantum electrodynamics and to be consistent with the expected UV divergence structure. Moreover, we impose the theoretical bounds from the perturbativity and the experimental constraints from LHC and LEP-II on our model. Especially, we derive novel perturbativity constraints on nonrenormalizable operators related to $G$, which are the natural generalization of the counterpart for the renormalizable operators. As a result, we show that there exists a substantial parameter space, which can accommodate the muon $g-2$ anomaly allowed by all constraints. Finally, we also make comments on the possible explanation of the electron $g-2$ anomalies with the massive spin-2 particle.

hep-ph↗