arXiv subjects

Wei Mao

Publications and source records attributed to Wei Mao.

36 records · Page 2Linked to original sources

arXiv2023

DeepSimHO: Stable Pose Estimation for Hand-Object Interaction via Physics Simulation

This paper addresses the task of 3D pose estimation for a hand interacting with an object from a single image observation. When modeling hand-object interaction, previous works mainly exploit proximity cues, while overlooking the dynamical nature that the hand must stably grasp the object to counteract gravity and thus preventing the object from slipping or falling. These works fail to leverage dynamical constraints in the estimation and consequently often produce unstable results. Meanwhile, refining unstable configurations with physics-based reasoning remains challenging, both by the complexity of contact dynamics and by the lack of effective and efficient physics inference in the data-driven learning framework. To address both issues, we present DeepSimHO: a novel deep-learning pipeline that combines forward physics simulation and backward gradient approximation with a neural network. Specifically, for an initial hand-object pose estimated by a base network, we forward it to a physics simulator to evaluate its stability. However, due to non-smooth contact geometry and penetration, existing differentiable simulators can not provide reliable state gradient. To remedy this, we further introduce a deep network to learn the stability evaluation process from the simulator, while smoothly approximating its gradient and thus enabling effective back-propagation. Extensive experiments show that our method noticeably improves the stability of the estimation and achieves superior efficiency over test-time optimization. The code is available at https://github.com/rongakowang/DeepSimHO.

cs.CV↗

arXiv2023

4D-Explorer: A visual software for 4D-STEM data processing and image reconstruction

With the development of high-speed electron detectors, four-dimensional scanning transmission electron microscopy (4D-STEM) has emerged as a powerful tool for characterizing microstructures in material science and life science. However, the complexity of 4D-STEM data processing necessitates an intuitive graphical user interface software for researchers. In this regard, we have developed 4D-Explorer, an open-source, lightweight and extensible software for processing 4D-STEM data. It offers a visual and interactive workflow, including data preparation, calibration, image reconstruction and generating quantitative results. Furthermore, during calibration, our software includes a novel algorithm for rotational offset correction that uses a defocused 4D-STEM dataset and its axial bright field image, which has lower experimental requirements than conventional methods. We anticipate that 4D-Explorer will help researchers harness the capabilities of 4D-STEM technology.

physics.ins-det↗

arXiv2023

VisFusion: Visibility-aware Online 3D Scene Reconstruction from Videos

We propose VisFusion, a visibility-aware online 3D scene reconstruction approach from posed monocular videos. In particular, we aim to reconstruct the scene from volumetric features. Unlike previous reconstruction methods which aggregate features for each voxel from input views without considering its visibility, we aim to improve the feature fusion by explicitly inferring its visibility from a similarity matrix, computed from its projected features in each image pair. Following previous works, our model is a coarse-to-fine pipeline including a volume sparsification process. Different from their works which sparsify voxels globally with a fixed occupancy threshold, we perform the sparsification on a local feature volume along each visual ray to preserve at least one voxel per ray for more fine details. The sparse local volume is then fused with a global one for online reconstruction. We further propose to predict TSDF in a coarse-to-fine manner by learning its residuals across scales leading to better TSDF predictions. Experimental results on benchmarks show that our method can achieve superior performance with more scene details. Code is available at: https://github.com/huiyu-gao/VisFusion

cs.CV↗

arXiv2022

Interacting Hand-Object Pose Estimation via Dense Mutual Attention

3D hand-object pose estimation is the key to the success of many computer vision applications. The main focus of this task is to effectively model the interaction between the hand and an object. To this end, existing works either rely on interaction constraints in a computationally-expensive iterative optimization, or consider only a sparse correlation between sampled hand and object keypoints. In contrast, we propose a novel dense mutual attention mechanism that is able to model fine-grained dependencies between the hand and the object. Specifically, we first construct the hand and object graphs according to their mesh structures. For each hand node, we aggregate features from every object node by the learned attention and vice versa for each object node. Thanks to such dense mutual attention, our method is able to produce physically plausible poses with high quality and real-time inference speed. Extensive quantitative and qualitative experiments on large benchmark datasets show that our method outperforms state-of-the-art methods. The code is available at https://github.com/rongakowang/DenseMutualAttention.git.

cs.CV↗

arXiv2022

Contact-aware Human Motion Forecasting

In this paper, we tackle the task of scene-aware 3D human motion forecasting, which consists of predicting future human poses given a 3D scene and a past human motion. A key challenge of this task is to ensure consistency between the human and the scene, accounting for human-scene interactions. Previous attempts to do so model such interactions only implicitly, and thus tend to produce artifacts such as "ghost motion" because of the lack of explicit constraints between the local poses and the global motion. Here, by contrast, we propose to explicitly model the human-scene contacts. To this end, we introduce distance-based contact maps that capture the contact relationships between every joint and every 3D scene point at each time instant. We then develop a two-stage pipeline that first predicts the future contact maps from the past ones and the scene point cloud, and then forecasts the future human poses by conditioning them on the predicted contact maps. During training, we explicitly encourage consistency between the global motion and the local poses via a prior defined using the contact maps and future poses. Our approach outperforms the state-of-the-art human motion forecasting and human synthesis methods on both synthetic and real datasets. Our code is available at https://github.com/wei-mao-2019/ContAwareMotionPred.

cs.CV↗

arXiv2022

Weakly-supervised Action Transition Learning for Stochastic Human Motion Prediction

We introduce the task of action-driven stochastic human motion prediction, which aims to predict multiple plausible future motions given a sequence of action labels and a short motion history. This differs from existing works, which predict motions that either do not respect any specific action category, or follow a single action label. In particular, addressing this task requires tackling two challenges: The transitions between the different actions must be smooth; the length of the predicted motion depends on the action sequence and varies significantly across samples. As we cannot realistically expect training data to cover sufficiently diverse action transitions and motion lengths, we propose an effective training strategy consisting of combining multiple motions from different actions and introducing a weak form of supervision to encourage smooth transitions. We then design a VAE-based model conditioned on both the observed motion and the action label sequence, allowing us to generate multiple plausible future motions of varying length. We illustrate the generality of our approach by exploring its use with two different temporal encoding models, namely RNNs and Transformers. Our approach outperforms baseline models constructed by adapting state-of-the-art single action-conditioned motion generation methods and stochastic human motion prediction approaches to our new task of action-driven stochastic motion prediction. Our code is available at https://github.com/wei-mao-2019/WAT.

cs.CV↗

arXiv2022

Integrated Access and Backhaul in Millimeter-Wave Cellular: Benefits and Challenges

The recently proposed NR-ready integrated access and backhaul (IAB) architecture promises to bring a cost-efficient deployment solution for both coverage extension and capacity boosting in future 5G/5G+ systems. While its impact on the coverage extension was thoughtfully addressed in the literature, the effect of advanced functionalities such as multi-hop, multi-connectivity, and multi-beam operations on the throughput remains unclear. We review and characterize the system-level impact of these capabilities on the performance of self-backhauled IAB systems operating in half-duplex mode and utilizing millimeter-wave (mmWave) technology across both access and backhaul. Our results indicate that the throughput gain of multi-hopping and multi-beaming is significant even without multi-connectivity operation. Another important learning is that in all-mmWave systems with link blockage, multi-connectivity with link switching allows achieving self-load balancing. Finally, we outline future research directions.

cs.NI↗

arXiv2022

Generating Smooth Pose Sequences for Diverse Human Motion Prediction

Recent progress in stochastic motion prediction, i.e., predicting multiple possible future human motions given a single past pose sequence, has led to producing truly diverse future motions and even providing control over the motion of some body parts. However, to achieve this, the state-of-the-art method requires learning several mappings for diversity and a dedicated model for controllable motion prediction. In this paper, we introduce a unified deep generative network for both diverse and controllable motion prediction. To this end, we leverage the intuition that realistic human motions consist of smooth sequences of valid poses, and that, given limited data, learning a pose prior is much more tractable than a motion one. We therefore design a generator that predicts the motion of different body parts sequentially, and introduce a normalizing flow based pose prior, together with a joint angle loss, to achieve motion realism.Our experiments on two standard benchmark datasets, Human3.6M and HumanEva-I, demonstrate that our approach outperforms the state-of-the-art baselines in terms of both sample diversity and accuracy. The code is available at https://github.com/wei-mao-2019/gsps

cs.CV↗

arXiv2021

Multi-level Motion Attention for Human Motion Prediction

Human motion prediction aims to forecast future human poses given a historical motion. Whether based on recurrent or feed-forward neural networks, existing learning based methods fail to model the observation that human motion tends to repeat itself, even for complex sports actions and cooking activities. Here, we introduce an attention based feed-forward network that explicitly leverages this observation. In particular, instead of modeling frame-wise attention via pose similarity, we propose to extract motion attention to capture the similarity between the current motion context and the historical motion sub-sequences. In this context, we study the use of different types of attention, computed at joint, body part, and full pose levels. Aggregating the relevant past motions and processing the result with a graph convolutional network allows us to effectively exploit motion patterns from the long-term history to predict the future poses. Our experiments on Human3.6M, AMASS and 3DPW validate the benefits of our approach for both periodical and non-periodical actions. Thanks to our attention model, it yields state-of-the-art results on all three datasets. Our code is available at https://github.com/wei-mao-2019/HisRepItself.

cs.CV↗

arXiv2021

Panoptic Lintention Network: Towards Efficient Navigational Perception for the Visually Impaired

Classic computer vision algorithms, instance segmentation, and semantic segmentation can not provide a holistic understanding of the surroundings for the visually impaired. In this paper, we utilize panoptic segmentation to assist the navigation of visually impaired people by offering both things and stuff awareness in the proximity of the visually impaired efficiently. To this end, we propose an efficient Attention module -- Lintention which can model long-range interactions in linear time using linear space. Based on Lintention, we then devise a novel panoptic segmentation model which we term Panoptic Lintention Net. Experiments on the COCO dataset indicate that the Panoptic Lintention Net raises the Panoptic Quality (PQ) from 39.39 to 41.42 with 4.6\% performance gain while only requiring 10\% fewer GFLOPs and 25\% fewer parameters in the semantic branch. Furthermore, a real-world test via our designed compact wearable panoptic segmentation system, indicates that our system based on the Panoptic Lintention Net accomplishes a relatively stable and exceptionally remarkable panoptic segmentation in real-world scenes.

cs.CV↗

arXiv2020

History Repeats Itself: Human Motion Prediction via Motion Attention

Human motion prediction aims to forecast future human poses given a past motion. Whether based on recurrent or feed-forward neural networks, existing methods fail to model the observation that human motion tends to repeat itself, even for complex sports actions and cooking activities. Here, we introduce an attention-based feed-forward network that explicitly leverages this observation. In particular, instead of modeling frame-wise attention via pose similarity, we propose to extract motion attention to capture the similarity between the current motion context and the historical motion sub-sequences. Aggregating the relevant past motions and processing the result with a graph convolutional network allows us to effectively exploit motion patterns from the long-term history to predict the future poses. Our experiments on Human3.6M, AMASS and 3DPW evidence the benefits of our approach for both periodical and non-periodical actions. Thanks to our attention model, it yields state-of-the-art results on all three datasets. Our code is available at https://github.com/wei-mao-2019/HisRepItself.

cs.CV↗

arXiv2020

Can we cover navigational perception needs of the visually impaired by panoptic segmentation?

Navigational perception for visually impaired people has been substantially promoted by both classic and deep learning based segmentation methods. In classic visual recognition methods, the segmentation models are mostly object-dependent, which means a specific algorithm has to be devised for the object of interest. In contrast, deep learning based models such as instance segmentation and semantic segmentation allow to individually recognize part of the entire scene, namely things or stuff, for blind individuals. However, both of them can not provide a holistic understanding of the surroundings for the visually impaired. Panoptic segmentation is a newly proposed visual model with the aim of unifying semantic segmentation and instance segmentation. Motivated by that, we propose to utilize panoptic segmentation as an approach to navigating visually impaired people by offering both things and stuff awareness in the proximity of the visually impaired. We demonstrate that panoptic segmentation is able to equip the visually impaired with a holistic real-world scene perception through a wearable assistive system.

cs.CV↗

arXiv2020

Learning Trajectory Dependencies for Human Motion Prediction

Human motion prediction, i.e., forecasting future body poses given observed pose sequence, has typically been tackled with recurrent neural networks (RNNs). However, as evidenced by prior work, the resulted RNN models suffer from prediction errors accumulation, leading to undesired discontinuities in motion prediction. In this paper, we propose a simple feed-forward deep network for motion prediction, which takes into account both temporal smoothness and spatial dependencies among human body joints. In this context, we then propose to encode temporal information by working in trajectory space, instead of the traditionally-used pose space. This alleviates us from manually defining the range of temporal dependencies (or temporal convolutional filter size, as done in previous work). Moreover, spatial dependency of human pose is encoded by treating a human pose as a generic graph (rather than a human skeletal kinematic tree) formed by links between every pair of body joints. Instead of using a pre-defined graph structure, we design a new graph convolutional network to learn graph connectivity automatically. This allows the network to capture long range dependencies beyond that of human kinematic tree. We evaluate our approach on several standard benchmark datasets for motion prediction, including Human3.6M, the CMU motion capture dataset and 3DPW. Our experiments clearly demonstrate that the proposed approach achieves state of the art performance, and is applicable to both angle-based and position-based pose representations. The code is available at https://github.com/wei-mao-2019/LearnTrajDep

cs.CV↗

arXiv2020

Cost Volume Pyramid Based Depth Inference for Multi-View Stereo

We propose a cost volume-based neural network for depth inference from multi-view images. We demonstrate that building a cost volume pyramid in a coarse-to-fine manner instead of constructing a cost volume at a fixed resolution leads to a compact, lightweight network and allows us inferring high resolution depth maps to achieve better reconstruction results. To this end, we first build a cost volume based on uniform sampling of fronto-parallel planes across the entire depth range at the coarsest resolution of an image. Then, given current depth estimate, we construct new cost volumes iteratively on the pixelwise depth residual to perform depth map refinement. While sharing similar insight with Point-MVSNet as predicting and refining depth iteratively, we show that working on cost volume pyramid can lead to a more compact, yet efficient network structure compared with the Point-MVSNet on 3D points. We further provide detailed analyses of the relation between (residual) depth sampling and image resolution, which serves as a principle for building compact cost volume pyramid. Experimental results on benchmark datasets show that our model can perform 6x faster and has similar performance as state-of-the-art methods. Code is available at https://github.com/JiayuYANG/CVP-MVSNet

cs.CV↗

arXiv2018

Models and information-theoretic bounds for nanopore sequencing

Nanopore sequencing is an emerging new technology for sequencing DNA, which can read long fragments of DNA (~50,000 bases) in contrast to most current short-read sequencing technologies which can only read hundreds of bases. While nanopore sequencers can acquire long reads, the high error rates (20%-30%) pose a technical challenge. In a nanopore sequencer, a DNA is migrated through a nanopore and current variations are measured. The DNA sequence is inferred from this observed current pattern using an algorithm called a base-caller. In this paper, we propose a mathematical model for the "channel" from the input DNA sequence to the observed current, and calculate bounds on the information extraction capacity of the nanopore sequencer. This model incorporates impairments like (non-linear) inter-symbol interference, deletions, as well as random response. These information bounds have two-fold application: (1) The decoding rate with a uniform input distribution can be used to calculate the average size of the plausible list of DNA sequences given an observed current trace. This bound can be used to benchmark existing base-calling algorithms, as well as serving a performance objective to design better nanopores. (2) When the nanopore sequencer is used as a reader in a DNA storage system, the storage capacity is quantified by our bounds.

cs.IT↗

arXiv2016

Capacity Analysis of Discrete Energy Harvesting Channels

We study the channel capacity of a general discrete energy harvesting channel with a finite battery. Contrary to traditional communication systems, the transmitter of such a channel is powered by a device that harvests energy from a random exogenous energy source and has a finite-sized battery. As a consequence, at each transmission opportunity the system can only transmit a symbol whose energy is no more than the energy currently available. This new type of power supply introduces an unprecedented input constraint for the channel, which is simultaneously random, instantaneous, and influenced by the full history of the inputs and the energy harvesting process. Furthermore, naturally, in such a channel the energy information is observed causally at the transmitter. Both of these characteristics pose great challenges for the analysis of the channel capacity. In this work we use techniques developed for channels with side information and finite state channels, to obtain lower and upper bounds on the capacity of energy harvesting channels. In particular, in a general case with Markov energy harvesting processes we use stationarity and ergodicity theory to compute and optimize the achievable rates for the channels, and derive series of computable capacity upper and lower bounds.

cs.IT↗

arXiv2014

On the Ingleton-Violations in Finite Groups

Given $n$ discrete random variables, its entropy vector is the $2^n-1$ dimensional vector obtained from the joint entropies of all non-empty subsets of the random variables. It is well known that there is a one-to-one correspondence between such an entropy vector and a certain group-characterizable vector obtained from a finite group and $n$ of its subgroups [3]. This correspondence may be useful for characterizing the space of entropic vectors and for designing network codes. If one restricts attention to abelian groups then not all entropy vectors can be obtained. This is an explanation for the fact shown by Dougherty et al [4] that linear network codes cannot achieve capacity in general network coding problems. All abelian group-characterizable vectors, and by fiat all entropy vectors generated by linear network codes, satisfy a linear inequality called the Ingleton inequality. It is therefore of interest to identify groups that violate the Ingleton inequality. In this paper, we study the problem of finding nonabelian finite groups that yield characterizable vectors which violate the Ingleton inequality. Using a refined computer search, we find the symmetric group $S_5$ to be the smallest group that violates the Ingleton inequality. Careful study of the structure of this group, and its subgroups, reveals that it belongs to the Ingleton-violating family $PGL(2,q)$ with a prime power $q \geq 5$, i.e., the projective group of $2\times 2$ nonsingular matrices with entries in $\mathbb{F}_q$. We further interpret this family using the theory of group actions. We also extend the construction to more general groups such as $PGL(n,q)$ and $GL(n,q)$. The families of groups identified here are therefore good candidates for constructing network codes more powerful than linear network codes, and we discuss some considerations for constructing such group network codes.

cs.IT↗

arXiv2009

Violating the Ingleton Inequality with Finite Groups

It is well known that there is a one-to-one correspondence between the entropy vector of a collection of n random variables and a certain group-characterizable vector obtained from a finite group and n of its subgroups. However, if one restricts attention to abelian groups then not all entropy vectors can be obtained. This is an explanation for the fact shown by Dougherty et al that linear network codes cannot achieve capacity in general network coding problems (since linear network codes form an abelian group). All abelian group-characterizable vectors, and by fiat all entropy vectors generated by linear network codes, satisfy a linear inequality called the Ingleton inequality. In this paper, we study the problem of finding nonabelian finite groups that yield characterizable vectors which violate the Ingleton inequality. Using a refined computer search, we find the symmetric group S_5 to be the smallest group that violates the Ingleton inequality. Careful study of the structure of this group, and its subgroups, reveals that it belongs to the Ingleton-violating family PGL(2,p) with primes p > 3, i.e., the projective group of 2 by 2 nonsingular matrices with entries in F_p. This family of groups is therefore a good candidate for constructing network codes more powerful than linear network codes.

cs.IT↗