arXiv subjects

Shen Zhao

Publications and source records attributed to Shen Zhao.

At least 37 records · Page 2Linked to original sources

arXiv2024

VidMan: Exploiting Implicit Dynamics from Video Diffusion Model for Effective Robot Manipulation

Recent advancements utilizing large-scale video data for learning video generation models demonstrate significant potential in understanding complex physical dynamics. It suggests the feasibility of leveraging diverse robot trajectory data to develop a unified, dynamics-aware model to enhance robot manipulation. However, given the relatively small amount of available robot data, directly fitting data without considering the relationship between visual observations and actions could lead to suboptimal data utilization. To this end, we propose VidMan (Video Diffusion for Robot Manipulation), a novel framework that employs a two-stage training mechanism inspired by dual-process theory from neuroscience to enhance stability and improve data utilization efficiency. Specifically, in the first stage, VidMan is pre-trained on the Open X-Embodiment dataset (OXE) for predicting future visual trajectories in a video denoising diffusion manner, enabling the model to develop a long horizontal awareness of the environment's dynamics. In the second stage, a flexible yet effective layer-wise self-attention adapter is introduced to transform VidMan into an efficient inverse dynamics model that predicts action modulated by the implicit dynamics knowledge via parameter sharing. Our VidMan framework outperforms state-of-the-art baseline model GR-1 on the CALVIN benchmark, achieving a 11.7% relative improvement, and demonstrates over 9% precision gains on the OXE small-scale dataset. These results provide compelling evidence that world models can significantly enhance the precision of robot action prediction. Codes and models will be public.

cs.CV↗

arXiv2024

Time-Delayed Koopman Network-Based Model Predictive Control for the FRIB RFQ

The radio-frequency quadrupole (RFQ) at the Facility for Rare Isotope Beams (FRIB) is a critical device to accelerate heavy ion beams from 12 keV/u to 0.5 MeV/u for state-of-the-art nuclear physics experiments. Efficient control of the RFQ resonance frequency detuning still remains a challenge because the temperature-sensitive frequency is solely control by a cooling water system, exhibiting complicated transport delay and nonlinearity in the heat transfer processes. In this work, we propose a long-short term memory (LSTM)-based Koopman network model that can simultaneously learn the time-delayed and non-delayed correlations hidden in the historical operating data. It is proven that the model can effectively predict the behavior of the RFQ resonance frequency using historical data as inputs. With this model, a model predictive control (MPC) framework based on the Newton-Raphson method is proposed and tested. We demonstrate that the MPC framework utilizing deep learning model is able to provide precise and rapid control for the RFQ frequency detuning, reducing the control time by half compared to the proportional-integral-derivative (PID) controller.

physics.acc-ph↗

arXiv2024

A Population-to-individual Tuning Framework for Adapting Pretrained LM to On-device User Intent Prediction

Mobile devices, especially smartphones, can support rich functions and have developed into indispensable tools in daily life. With the rise of generative AI services, smartphones can potentially transform into personalized assistants, anticipating user needs and scheduling services accordingly. Predicting user intents on smartphones, and reflecting anticipated activities based on past interactions and context, remains a pivotal step towards this vision. Existing research predominantly focuses on specific domains, neglecting the challenge of modeling diverse event sequences across dynamic contexts. Leveraging pre-trained language models (PLMs) offers a promising avenue, yet adapting PLMs to on-device user intent prediction presents significant challenges. To address these challenges, we propose PITuning, a Population-to-Individual Tuning framework. PITuning enhances common pattern extraction through dynamic event-to-intent transition modeling and addresses long-tailed preferences via adaptive unlearning strategies. Experimental results on real-world datasets demonstrate PITuning's superior intent prediction performance, highlighting its ability to capture long-tailed preferences and its practicality for on-device prediction scenarios.

cs.LG↗

arXiv2024

Predicting Genetic Mutation from Whole Slide Images via Biomedical-Linguistic Knowledge Enhanced Multi-label Classification

Predicting genetic mutations from whole slide images is indispensable for cancer diagnosis. However, existing work training multiple binary classification models faces two challenges: (a) Training multiple binary classifiers is inefficient and would inevitably lead to a class imbalance problem. (b) The biological relationships among genes are overlooked, which limits the prediction performance. To tackle these challenges, we innovatively design a Biological-knowledge enhanced PathGenomic multi-label Transformer to improve genetic mutation prediction performances. BPGT first establishes a novel gene encoder that constructs gene priors by two carefully designed modules: (a) A gene graph whose node features are the genes' linguistic descriptions and the cancer phenotype, with edges modeled by genes' pathway associations and mutation consistencies. (b) A knowledge association module that fuses linguistic and biomedical knowledge into gene priors by transformer-based graph representation learning, capturing the intrinsic relationships between different genes' mutations. BPGT then designs a label decoder that finally performs genetic mutation prediction by two tailored modules: (a) A modality fusion module that firstly fuses the gene priors with critical regions in WSIs and obtains gene-wise mutation logits. (b) A comparative multi-label loss that emphasizes the inherent comparisons among mutation status to enhance the discrimination capabilities. Sufficient experiments on The Cancer Genome Atlas benchmark demonstrate that BPGT outperforms the state-of-the-art.

cs.CV↗

arXiv2024

VG4D: Vision-Language Model Goes 4D Video Recognition

Understanding the real world through point cloud video is a crucial aspect of robotics and autonomous driving systems. However, prevailing methods for 4D point cloud recognition have limitations due to sensor resolution, which leads to a lack of detailed information. Recent advances have shown that Vision-Language Models (VLM) pre-trained on web-scale text-image datasets can learn fine-grained visual concepts that can be transferred to various downstream tasks. However, effectively integrating VLM into the domain of 4D point clouds remains an unresolved problem. In this work, we propose the Vision-Language Models Goes 4D (VG4D) framework to transfer VLM knowledge from visual-text pre-trained models to a 4D point cloud network. Our approach involves aligning the 4D encoder's representation with a VLM to learn a shared visual and text space from training on large-scale image-text pairs. By transferring the knowledge of the VLM to the 4D encoder and combining the VLM, our VG4D achieves improved recognition performance. To enhance the 4D encoder, we modernize the classic dynamic point cloud backbone and propose an improved version of PSTNet, im-PSTNet, which can efficiently model point cloud videos. Experiments demonstrate that our method achieves state-of-the-art performance for action recognition on both the NTU RGB+D 60 dataset and the NTU RGB+D 120 dataset. Code is available at \url{https://github.com/Shark0-0/VG4D}.

cs.CV↗

arXiv2024

Hybrid moiré excitons and trions in twisted MoTe$_2$-MoSe$_2$ heterobilayers

We report experimental and theoretical studies of MoTe$_2$-MoSe$_2$ heterobilayers with rigid moiré superlattices controlled by the twist angle. Using an effective continuum model that combines resonant interlayer electron tunneling with stacking-dependent moiré potentials, we identify the nature of moiré excitons and the dependence of their energies, oscillator strengths and Landé $g$-factors on the twist angle. Within the same framework, we interpret distinct signatures of bound complexes among electrons and moiré excitons in nearly collinear heterostacks. Our work provides fundamental understanding of hybrid moiré excitons and trions in MoTe$_2$-MoSe$_2$ heterobilayers, and establishes the material system as a prime candidate for optical studies of correlated phenomena in moiré lattices.

cond-mat.mes-hall↗

arXiv2024

ModelNet-O: A Large-Scale Synthetic Dataset for Occlusion-Aware Point Cloud Classification

Recently, 3D point cloud classification has made significant progress with the help of many datasets. However, these datasets do not reflect the incomplete nature of real-world point clouds caused by occlusion, which limits the practical application of current methods. To bridge this gap, we propose ModelNet-O, a large-scale synthetic dataset of 123,041 samples that emulate real-world point clouds with self-occlusion caused by scanning from monocular cameras. ModelNet-O is 10 times larger than existing datasets and offers more challenging cases to evaluate the robustness of existing methods. Our observation on ModelNet-O reveals that well-designed sparse structures can preserve structural information of point clouds under occlusion, motivating us to propose a robust point cloud processing method that leverages a critical point sampling (CPS) strategy in a multi-level manner. We term our method PointMLS. Through extensive experiments, we demonstrate that our PointMLS achieves state-of-the-art results on ModelNet-O and competitive results on regular datasets, and it is robust and effective. More experiments also demonstrate the robustness and effectiveness of PointMLS.

cs.CV↗

arXiv2024

Explore Human Parsing Modality for Action Recognition

Multimodal-based action recognition methods have achieved high success using pose and RGB modality. However, skeletons sequences lack appearance depiction and RGB images suffer irrelevant noise due to modality limitations. To address this, we introduce human parsing feature map as a novel modality, since it can selectively retain effective semantic features of the body parts, while filtering out most irrelevant noise. We propose a new dual-branch framework called Ensemble Human Parsing and Pose Network (EPP-Net), which is the first to leverage both skeletons and human parsing modalities for action recognition. The first human pose branch feeds robust skeletons in graph convolutional network to model pose features, while the second human parsing branch also leverages depictive parsing feature maps to model parsing festures via convolutional backbones. The two high-level features will be effectively combined through a late fusion strategy for better action recognition. Extensive experiments on NTU RGB+D and NTU RGB+D 120 benchmarks consistently verify the effectiveness of our proposed EPP-Net, which outperforms the existing action recognition methods. Our code is available at: https://github.com/liujf69/EPP-Net-Action.

cs.CV↗

arXiv2024

Precision-controlled ultrafast electron microscope platforms. A case study: Multiple-order coherent phonon dynamics in 1T-TaSe$_2$ probed at 50 femtosecond - 10 femtometer scales

We report on the first detailed beam test attesting the fundamental principle behind the development of high-current-efficiency ultrafast electron microscope systems where a radio-frequency cavity is incorporated as a condenser lens in the beam delivery system. To allow the experiment to be carried out with a sufficient resolution to probe the performance at the emittance floor, a new cascade loop RF controller system is developed to reduce the RF noise floor. Temporal resolution at 50 femtoseconds in full-width-at-half-maximum and detection sensitivity better than 1% are demonstrated on exfoliated 1T-TaSe$_2$ layers where the multi-order edge-mode coherent phonon excitation is employed as the standard candle to benchmark the performance. The high temporal resolution and the significant visibility to very low dynamical contrast in diffraction signals give strong support to the working principle of the high-brightness beam delivery via phase-space manipulation in the electron microscope system.

cond-mat.mtrl-sci↗

arXiv2023

Dynamic Compositional Graph Convolutional Network for Efficient Composite Human Motion Prediction

With potential applications in fields including intelligent surveillance and human-robot interaction, the human motion prediction task has become a hot research topic and also has achieved high success, especially using the recent Graph Convolutional Network (GCN). Current human motion prediction task usually focuses on predicting human motions for atomic actions. Observing that atomic actions can happen at the same time and thus formulating the composite actions, we propose the composite human motion prediction task. To handle this task, we first present a Composite Action Generation (CAG) module to generate synthetic composite actions for training, thus avoiding the laborious work of collecting composite action samples. Moreover, we alleviate the effect of composite actions on demand for a more complicated model by presenting a Dynamic Compositional Graph Convolutional Network (DC-GCN). Extensive experiments on the Human3.6M dataset and our newly collected CHAMP dataset consistently verify the efficiency of our DC-GCN method, which achieves state-of-the-art motion prediction accuracies and meanwhile needs few extra computational costs than traditional GCN-based human motion methods.

cs.CV↗

arXiv2023

Exciton-exciton interactions in van der Waals heterobilayers

Exciton-exciton interactions are key to understanding non-linear optical and transport phenomena in van der Waals heterobilayers, which emerged as versatile platforms to study correlated electronic states. We present a combined theory-experiment study of excitonic many-body effects based on first-principle band structures and Coulomb interaction matrix elements. Key to our approach is the explicit treatment of the fermionic substructure of excitons and dynamical screening effects for density-induced energy renormalization and dissipation. We demonstrate that dipolar blue shifts are almost perfectly compensated by many-body effects, mainly by screening-induced self-energy corrections. Moreover, we identify a crossover between attractive and repulsive behavior at elevated exciton densities. Theoretical findings are supported by experimental studies of spectrally-narrow interlayer excitons in atomically-reconstructed, hBN-encapsulated MoSe$_2$/WSe$_2$ heterobilayers. Both theory and experiment show energy renormalization on a scale of a few meV even for high injection densities in the vicinity of the Mott transition. Our results revise the established picture of dipolar repulsion dominating exciton-exciton interactions in van der Waals heterostructures and open up opportunities for their external design.

cond-mat.mes-hall↗

arXiv2023

GrowCLIP: Data-aware Automatic Model Growing for Large-scale Contrastive Language-Image Pre-training

Cross-modal pre-training has shown impressive performance on a wide range of downstream tasks, benefiting from massive image-text pairs collected from the Internet. In practice, online data are growing constantly, highlighting the importance of the ability of pre-trained model to learn from data that is continuously growing. Existing works on cross-modal pre-training mainly focus on training a network with fixed architecture. However, it is impractical to limit the model capacity when considering the continuously growing nature of pre-training data in real-world applications. On the other hand, it is important to utilize the knowledge in the current model to obtain efficient training and better performance. To address the above issues, in this paper, we propose GrowCLIP, a data-driven automatic model growing algorithm for contrastive language-image pre-training with continuous image-text pairs as input. Specially, we adopt a dynamic growth space and seek out the optimal architecture at each growth step to adapt to online learning scenarios. And the shared encoder is proposed in our growth space to enhance the degree of cross-modal fusion. Besides, we explore the effect of growth in different dimensions, which could provide future references for the design of cross-modal model architecture. Finally, we employ parameter inheriting with momentum (PIM) to maintain the previous knowledge and address the issue of the local minimum dilemma. Compared with the existing methods, GrowCLIP improves 2.3% average top-1 accuracy on zero-shot image classification of 9 downstream tasks. As for zero-shot image retrieval, GrowCLIP can improve 1.2% for top-1 image-to-text recall on Flickr30K dataset.

cs.CV↗

arXiv2023

Graph-Guided MLP-Mixer for Skeleton-Based Human Motion Prediction

In recent years, Graph Convolutional Networks (GCNs) have been widely used in human motion prediction, but their performance remains unsatisfactory. Recently, MLP-Mixer, initially developed for vision tasks, has been leveraged into human motion prediction as a promising alternative to GCNs, which achieves both better performance and better efficiency than GCNs. Unlike GCNs, which can explicitly capture human skeleton's bone-joint structure by representing it as a graph with edges and nodes, MLP-Mixer relies on fully connected layers and thus cannot explicitly model such graph-like structure of human's. To break this limitation of MLP-Mixer's, we propose \textit{Graph-Guided Mixer}, a novel approach that equips the original MLP-Mixer architecture with the capability to model graph structure. By incorporating graph guidance, our \textit{Graph-Guided Mixer} can effectively capture and utilize the specific connectivity patterns within human skeleton's graph representation. In this paper, first we uncover a theoretical connection between MLP-Mixer and GCN that is unexplored in existing research. Building on this theoretical connection, next we present our proposed \textit{Graph-Guided Mixer}, explaining how the original MLP-Mixer architecture is reinvented to incorporate guidance from graph structure. Then we conduct an extensive evaluation on the Human3.6M, AMASS, and 3DPW datasets, which shows that our method achieves state-of-the-art performance.

cs.CV↗

arXiv2023

Learning Snippet-to-Motion Progression for Skeleton-based Human Motion Prediction

Existing Graph Convolutional Networks to achieve human motion prediction largely adopt a one-step scheme, which output the prediction straight from history input, failing to exploit human motion patterns. We observe that human motions have transitional patterns and can be split into snippets representative of each transition. Each snippet can be reconstructed from its starting and ending poses referred to as the transitional poses. We propose a snippet-to-motion multi-stage framework that breaks motion prediction into sub-tasks easier to accomplish. Each sub-task integrates three modules: transitional pose prediction, snippet reconstruction, and snippet-to-motion prediction. Specifically, we propose to first predict only the transitional poses. Then we use them to reconstruct the corresponding snippets, obtaining a close approximation to the true motion sequence. Finally we refine them to produce the final prediction output. To implement the network, we propose a novel unified graph modeling, which allows for direct and effective feature propagation compared to existing approaches which rely on separate space-time modeling. Extensive experiments on Human 3.6M, CMU Mocap and 3DPW datasets verify the effectiveness of our method which achieves state-of-the-art performance.

cs.CV↗

arXiv2023

Lattice reconstruction in MoSe$_2$-WSe$_2$ heterobilayers synthesized by chemical vapor deposition

Vertical van der Waals heterostructures of semiconducting transition metal dichalcogenides realize moiré systems with rich correlated electron phases and moiré exciton phenomena. For material combinations with small lattice mismatch and twist angles as in MoSe$_2$-WSe$_2$, however, lattice reconstruction eliminates the canonical moiré pattern and instead gives rise to arrays of periodically reconstructed nanoscale domains and mesoscopically extended areas of one atomic registry. Here, we elucidate the role of atomic reconstruction in MoSe$_2$-WSe$_2$ heterostructures synthesized by chemical vapor deposition. With complementary imaging down to the atomic scale, simulations, and optical spectroscopy methods we identify the coexistence of moiré-type cores and extended moiré-free regions in heterostacks with parallel and antiparallel alignment. Our work highlights the potential of chemical vapor deposition for applications requiring laterally extended heterosystems of one atomic registry or exciton-confining heterostack arrays.

cond-mat.mes-hall↗

arXiv2023

Towards Medical Artificial General Intelligence via Knowledge-Enhanced Multimodal Pretraining

Medical artificial general intelligence (MAGI) enables one foundation model to solve different medical tasks, which is very practical in the medical domain. It can significantly reduce the requirement of large amounts of task-specific data by sufficiently sharing medical knowledge among different tasks. However, due to the challenges of designing strongly generalizable models with limited and complex medical data, most existing approaches tend to develop task-specific models. To take a step towards MAGI, we propose a new paradigm called Medical-knOwledge-enhanced mulTimOdal pretRaining (MOTOR). In MOTOR, we combine two kinds of basic medical knowledge, i.e., general and specific knowledge, in a complementary manner to boost the general pretraining process. As a result, the foundation model with comprehensive basic knowledge can learn compact representations from pretraining radiographic data for better cross-modal alignment. MOTOR unifies the understanding and generation, which are two kinds of core intelligence of an AI system, into a single medical foundation model, to flexibly handle more diverse medical tasks. To enable a comprehensive evaluation and facilitate further research, we construct a medical multimodal benchmark including a wide range of downstream tasks, such as chest x-ray report generation and medical visual question answering. Extensive experiments on our benchmark show that MOTOR obtains promising results through simple task-oriented adaptation. The visualization shows that the injected knowledge successfully highlights key information in the medical data, demonstrating the excellent interpretability of MOTOR. Our MOTOR successfully mimics the human practice of fulfilling a "medical student" to accelerate the process of becoming a "specialist". We believe that our work makes a significant stride in realizing MAGI.

cs.AI↗

arXiv2023

Near-Intrinsic Photo- and Electroluminescence from Single-Walled Carbon Nanotube Thin Films on BCB-Passivated Surfaces

Their outstanding electrical and optical properties make semiconducting single-walled carbon nanotubes (SWCNTs) highly suitable for charge transport and emissive layers in near-infrared optoelectronic devices. However, the luminescence spectra of SWCNT thin films on commonly used glass and Si/SiO2 substrates are often compromised by broadening of the main excitonic emission and unwanted low-energy sidebands. Surface passivation with a commercially available, low dielectric constant, cross-linked bis-benzocyclobutene-based polymer (BCB) enhances the emission properties of SWCNTs to the same level as hexagonal boron nitride (h-BN) flakes do. The presence of BCB suppresses sideband emission, especially from the Y1 band, which is attributed to defects introduced by the interaction of the nanotube lattice with oxygen-containing terminal groups of the substrate surface. The facile and reproducible deposition of homogeneous BCB films over large areas combined with their resistance against common solvents and chemicals employed during photolithography make them compatible with standard semiconductor device fabrication. Utilizing this approach, light-emitting (6,5) SWCNT network field-effect transistors are fabricated on BCB-treated glass substrates with excellent electrical characteristics and near-intrinsic electroluminescence. Hence, passivation with BCB is proposed as a standard treatment for substrates used for spectroscopic investigations of and optoelectronic devices with SWCNTs and other low-dimensional emitters.

physics.app-ph↗

arXiv2023

CapDet: Unifying Dense Captioning and Open-World Detection Pretraining

Benefiting from large-scale vision-language pre-training on image-text pairs, open-world detection methods have shown superior generalization ability under the zero-shot or few-shot detection settings. However, a pre-defined category space is still required during the inference stage of existing methods and only the objects belonging to that space will be predicted. To introduce a "real" open-world detector, in this paper, we propose a novel method named CapDet to either predict under a given category list or directly generate the category of predicted bounding boxes. Specifically, we unify the open-world detection and dense caption tasks into a single yet effective framework by introducing an additional dense captioning head to generate the region-grounded captions. Besides, adding the captioning task will in turn benefit the generalization of detection performance since the captioning dataset covers more concepts. Experiment results show that by unifying the dense caption task, our CapDet has obtained significant performance improvements (e.g., +2.1% mAP on LVIS rare classes) over the baseline method on LVIS (1203 classes). Besides, our CapDet also achieves state-of-the-art performance on dense captioning tasks, e.g., 15.44% mAP on VG V1.2 and 13.98% on the VG-COCO dataset.

cs.CV↗