arXiv subjects

Zhe Chen

Publications and source records attributed to Zhe Chen.

At least 235 records · Page 13Linked to original sources

arXiv2023

InternImage: Exploring Large-Scale Vision Foundation Models with Deformable Convolutions

Compared to the great progress of large-scale vision transformers (ViTs) in recent years, large-scale models based on convolutional neural networks (CNNs) are still in an early state. This work presents a new large-scale CNN-based foundation model, termed InternImage, which can obtain the gain from increasing parameters and training data like ViTs. Different from the recent CNNs that focus on large dense kernels, InternImage takes deformable convolution as the core operator, so that our model not only has the large effective receptive field required for downstream tasks such as detection and segmentation, but also has the adaptive spatial aggregation conditioned by input and task information. As a result, the proposed InternImage reduces the strict inductive bias of traditional CNNs and makes it possible to learn stronger and more robust patterns with large-scale parameters from massive data like ViTs. The effectiveness of our model is proven on challenging benchmarks including ImageNet, COCO, and ADE20K. It is worth mentioning that InternImage-H achieved a new record 65.4 mAP on COCO test-dev and 62.9 mIoU on ADE20K, outperforming current leading CNNs and ViTs. The code will be released at https://github.com/OpenGVLab/InternImage.

cs.CV↗

arXiv2023

FPGA-Based In-Vivo Calcium Image Decoding for Closed-Loop Feedback Applications

Miniaturized calcium imaging is an emerging neural recording technique that has been widely used for monitoring neural activity on a large scale at a specific brain region of rats or mice. Most existing calcium-image analysis pipelines operate offline. This results in long processing latency, making it difficult to realize closed-loop feedback stimulation for brain research. In recent work, we have proposed an FPGA-based real-time calcium image processing pipeline for closed-loop feedback applications. It can perform real-time calcium image motion correction, enhancement, fast trace extraction, and real-time decoding from extracted traces. Here, we extend this work by proposing a variety of neural network based methods for real-time decoding and evaluate the tradeoff among these decoding methods and accelerator designs. We introduced the implementation of the neural network based decoders on the FPGA, and showed their speedup against the implementation on the ARM processor. Our FPGA implementation enables the real-time calcium image decoding with sub-ms processing latency for closed-loop feedback applications.

eess.IV↗

arXiv2023

AutoFed: Heterogeneity-Aware Federated Multimodal Learning for Robust Autonomous Driving

Object detection with on-board sensors (e.g., lidar, radar, and camera) play a crucial role in autonomous driving (AD), and these sensors complement each other in modalities. While crowdsensing may potentially exploit these sensors (of huge quantity) to derive more comprehensive knowledge, \textit{federated learning} (FL) appears to be the necessary tool to reach this potential: it enables autonomous vehicles (AVs) to train machine learning models without explicitly sharing raw sensory data. However, the multimodal sensors introduce various data heterogeneity across distributed AVs (e.g., label quantity skews and varied modalities), posing critical challenges to effective FL. To this end, we present AutoFed as a heterogeneity-aware FL framework to fully exploit multimodal sensory data on AVs and thus enable robust AD. Specifically, we first propose a novel model leveraging pseudo-labeling to avoid mistakenly treating unlabeled objects as the background. We also propose an autoencoder-based data imputation method to fill missing data modality (of certain AVs) with the available ones. To further reconcile the heterogeneity, we finally present a client selection mechanism exploiting the similarities among client models to improve both training stability and convergence rate. Our experiments on benchmark dataset confirm that AutoFed substantially improves over status quo approaches in both precision and recall, while demonstrating strong robustness to adverse weather conditions.

cs.LG↗

arXiv2023

Multi-band Reconfigurable Holographic Surface Based ISAC Systems: Design and Optimization

Metamaterial-based reconfigurable holographic surfaces (RHSs) have been proposed as novel cost-efficient antenna arrays, which are promising for improving the positioning and communication performance of integrated sensing and communications (ISAC) systems. However, due to the high frequency selectivity of the metamaterial elements, RHSs face challenges in supporting ultra-wide bandwidth (UWB), which significantly limits the positioning precision. In this paper, to avoid the physical limitations of UWB RHS while enhancing the performance of RHS-based ISAC systems, we propose a multi-band (MB) based ISAC system. We analyze its positioning precision and propose an efficient algorithm to optimize the large number of variables in analog and digital beamforming. Through comparison with benchmark results, simulation results verify the efficiency of our proposed system and algorithm, and show that the system achieves $42\%$ less positioning error, which reduces $82\%$ communication capacity loss.

eess.SP↗

arXiv2023

Champion Solution for the WSDM2023 Toloka VQA Challenge

In this report, we present our champion solution to the WSDM2023 Toloka Visual Question Answering (VQA) Challenge. Different from the common VQA and visual grounding (VG) tasks, this challenge involves a more complex scenario, i.e. inferring and locating the object implicitly specified by the given interrogative question. For this task, we leverage ViT-Adapter, a pre-training-free adapter network, to adapt multi-modal pre-trained Uni-Perceiver for better cross-modal localization. Our method ranks first on the leaderboard, achieving 77.5 and 76.347 IoU on public and private test sets, respectively. It shows that ViT-Adapter is also an effective paradigm for adapting the unified perception model to vision-language downstream tasks. Code and models will be released at https://github.com/czczup/ViT-Adapter/tree/main/wsdm2023.

cs.CV↗

arXiv2023

A Fourier spectral immersed boundary method with exact translation invariance, improved boundary resolution, and a divergence-free velocity field

This paper introduces a new immersed boundary (IB) method for viscous incompressible flow, based on a Fourier spectral method for the fluid solver and on the nonuniform fast Fourier transform (NUFFT) algorithm for coupling the fluid with the immersed boundary. The new Fourier spectral immersed boundary (FSIB) method gives improved boundary resolution in comparison to the standard IB method. The interpolated velocity field, in which the boundary moves, is analytically divergence-free. The FSIB method is gridless and has the meritorious properties of volume conservation, exact translation invariance, conservation of momentum, and conservation of energy. We verify these advantages of the FSIB method numerically both for the Stokes equations and for the Navier-Stokes equations in both two and three space dimensions. The FSIB method converges faster than the IB method. In particular, we observe second-order convergence in various problems for the Navier-Stokes equations in three dimensions. The FSIB method is also computationally efficient with complexity of $O(N^3\log(N))$ per time step for $N^3$ Fourier modes in three dimensions.

physics.flu-dyn↗

arXiv2023

Vision Transformer Adapter for Dense Predictions

This work investigates a simple yet powerful dense prediction task adapter for Vision Transformer (ViT). Unlike recently advanced variants that incorporate vision-specific inductive biases into their architectures, the plain ViT suffers inferior performance on dense predictions due to weak prior assumptions. To address this issue, we propose the ViT-Adapter, which allows plain ViT to achieve comparable performance to vision-specific transformers. Specifically, the backbone in our framework is a plain ViT that can learn powerful representations from large-scale multi-modal data. When transferring to downstream tasks, a pre-training-free adapter is used to introduce the image-related inductive biases into the model, making it suitable for these tasks. We verify ViT-Adapter on multiple dense prediction tasks, including object detection, instance segmentation, and semantic segmentation. Notably, without using extra detection data, our ViT-Adapter-L yields state-of-the-art 60.9 box AP and 53.0 mask AP on COCO test-dev. We hope that the ViT-Adapter could serve as an alternative for vision-specific transformers and facilitate future research. The code and models will be released at https://github.com/czczup/ViT-Adapter.

cs.CV↗

arXiv2023

Cooperative trajectory planning algorithm of USV-UAV with hull dynamic constraints

Efficient trajectory generation in complex dynamic environments remains an open problem in the unmanned surface vehicle (USV). The perception of the USV is usually interfered with by the swing of the hull and the ambient weather, making it challenging to plan the optimal USV trajectories. In this paper, a cooperative trajectory planning algorithm for the coupled USV-UAV system is proposed to ensure that USV can execute a safe and smooth path in the process of autonomous advance in multi-obstacle maps. Specifically, the unmanned aerial vehicle (UAV) plays the role of a flight sensor, providing real-time global map and obstacle information with a lightweight semantic segmentation network and 3D projection transformation. And then, an initial obstacle avoidance trajectory is generated by a graph-based search method. Concerning the unique under-actuated kinematic characteristics of the USV, a numerical optimization method based on hull dynamic constraints is introduced to make the trajectory easier to be tracked for motion control. Finally, a motion control method based on NMPC with the lowest energy consumption constraint during execution is proposed. Experimental results verify the effectiveness of the whole system, and the generated trajectory is locally optimal for USV with considerable tracking accuracy.

cs.RO↗

arXiv2023

FAST: Faster Arbitrarily-Shaped Text Detector with Minimalist Kernel Representation

We propose an accurate and efficient scene text detection framework, termed FAST (i.e., faster arbitrarily-shaped text detector). Different from recent advanced text detectors that used complicated post-processing and hand-crafted network architectures, resulting in low inference speed, FAST has two new designs. (1) We design a minimalist kernel representation (only has 1-channel output) to model text with arbitrary shape, as well as a GPU-parallel post-processing to efficiently assemble text lines with a negligible time overhead. (2) We search the network architecture tailored for text detection, leading to more powerful features than most networks that are searched for image classification. Benefiting from these two designs, FAST achieves an excellent trade-off between accuracy and efficiency on several challenging datasets, including Total Text, CTW1500, ICDAR 2015, and MSRA-TD500. For example, FAST-T yields 81.6% F-measure at 152 FPS on Total-Text, outperforming the previous fastest method by 1.7 points and 70 FPS in terms of accuracy and speed. With TensorRT optimization, the inference speed can be further accelerated to over 600 FPS. Code and models will be released at https://github.com/czczup/FAST.

cs.CV↗

arXiv2022

LegalRelectra: Mixed-domain Language Modeling for Long-range Legal Text Comprehension

The application of Natural Language Processing (NLP) to specialized domains, such as the law, has recently received a surge of interest. As many legal services rely on processing and analyzing large collections of documents, automating such tasks with NLP tools emerges as a key challenge. Many popular language models, such as BERT or RoBERTa, are general-purpose models, which have limitations on processing specialized legal terminology and syntax. In addition, legal documents may contain specialized vocabulary from other domains, such as medical terminology in personal injury text. Here, we propose LegalRelectra, a legal-domain language model that is trained on mixed-domain legal and medical corpora. We show that our model improves over general-domain and single-domain medical and legal language models when processing mixed-domain (personal injury) text. Our training architecture implements the Electra framework, but utilizes Reformer instead of BERT for its generator and discriminator. We show that this improves the model's performance on processing long passages and results in better long-range text comprehension.

cs.CL↗

arXiv2022

InternVideo-Ego4D: A Pack of Champion Solutions to Ego4D Challenges

In this report, we present our champion solutions to five tracks at Ego4D challenge. We leverage our developed InternVideo, a video foundation model, for five Ego4D tasks, including Moment Queries, Natural Language Queries, Future Hand Prediction, State Change Object Detection, and Short-term Object Interaction Anticipation. InternVideo-Ego4D is an effective paradigm to adapt the strong foundation model to the downstream ego-centric video understanding tasks with simple head designs. In these five tasks, the performance of InternVideo-Ego4D comprehensively surpasses the baseline methods and the champions of CVPR2022, demonstrating the powerful representation ability of InternVideo as a video foundation model. Our code will be released at https://github.com/OpenGVLab/ego4d-eccv2022-solutions

cs.CV↗

arXiv2022

Efficient Trajectory Planning and Control for USV with Vessel Dynamics and Differential Flatness

Unmanned surface vessels (USVs) are widely used in ocean exploration and environmental protection fields. To ensure that USV can successfully perform its mission, trajectory planning and motion tracking are the two most critical technologies. In this paper, we propose a novel trajectory generation and tracking method for USV based on optimization theory. Specifically, the USV dynamic model is described with differential flatness, so that the trajectory can be generated by dynamic RRT* in a linear invariant system expression form under the objective of optimal boundary value. To reduce the sample number and improve efficiency, we adjust the trajectory through local optimization. The dynamic constraints are considered in the optimization process so that the generated trajectory conforms to the kinematic characteristics of the under-actuated hull, and makes it easier to be tracked. Finally, motion tracking is added with model predictive control under a sequential quadratic programming problem. Experimental results show the planned trajectory is more in line with the kinematic characteristics of USV, and the tracking accuracy remains a higher level.

cs.RO↗

arXiv2022

Beyond Greedy Search: Tracking by Multi-Agent Reinforcement Learning-based Beam Search

To track the target in a video, current visual trackers usually adopt greedy search for target object localization in each frame, that is, the candidate region with the maximum response score will be selected as the tracking result of each frame. However, we found that this may be not an optimal choice, especially when encountering challenging tracking scenarios such as heavy occlusion and fast motion. To address this issue, we propose to maintain multiple tracking trajectories and apply beam search strategy for visual tracking, so that the trajectory with fewer accumulated errors can be identified. Accordingly, this paper introduces a novel multi-agent reinforcement learning based beam search tracking strategy, termed BeamTracking. It is mainly inspired by the image captioning task, which takes an image as input and generates diverse descriptions using beam search algorithm. Accordingly, we formulate the tracking as a sample selection problem fulfilled by multiple parallel decision-making processes, each of which aims at picking out one sample as their tracking result in each frame. Each maintained trajectory is associated with an agent to perform the decision-making and determine what actions should be taken to update related information. When all the frames are processed, we select the trajectory with the maximum accumulated score as the tracking result. Extensive experiments on seven popular tracking benchmark datasets validated the effectiveness of the proposed algorithm.

cs.CV↗

arXiv2022

Design of a $μ$SR spectrometer with ultrahigh arrays at China Spallation Neutron Source

A new muon source has been proposed to conduct muon spin rotation/relaxation/resonance ($μ$SR) measurements at China Spallation Neutron Source (CSNS). Only 1 Hz of the CSNS proton beams (25 Hz in total) will be allocated for muon production. To make better use of muons in every pulse, an ultrahigh-array $μ$SR spectrometer (UHAM) with thousands of detector channels is under design. Due to such a high granularity of detectors, multiple counting events generated from particle scattering or spiral motions of positrons in a strong longitudinal field should be carefully considered in the design. Six different structures were modeled and simulated based on two types of angular orientations (parallel arrangement and pointing arrangement) and three kinds of spectrometer geometries (cylinder, cone and sphere). A quality factor, $Q$, was proposed to evaluate the performance of these structures by integrating their impacts on the overall asymmetry, the counting rate and the percentage of multiple counts. According to the simulation, the conical structure with detectors pointing to the sample has the highest $Q$ in both zero field and longitudinal field. The results also show that all kinds of structures cannot be operated under strong longitudinal fields with a strength over 2 T. The full simulation of a $μ$SR spectrometer can provide good guidance for the construction of the UHAM in the upcoming upgrade of CSNS.

physics.ins-det↗

arXiv2022

Transformer-based Context Condensation for Boosting Feature Pyramids in Object Detection

Current object detectors typically have a feature pyramid (FP) module for multi-level feature fusion (MFF) which aims to mitigate the gap between features from different levels and form a comprehensive object representation to achieve better detection performance. However, they usually require heavy cross-level connections or iterative refinement to obtain better MFF result, making them complicated in structure and inefficient in computation. To address these issues, we propose a novel and efficient context modeling mechanism that can help existing FPs deliver better MFF results while reducing the computational costs effectively. In particular, we introduce a novel insight that comprehensive contexts can be decomposed and condensed into two types of representations for higher efficiency. The two representations include a locally concentrated representation and a globally summarized representation, where the former focuses on extracting context cues from nearby areas while the latter extracts key representations of the whole image scene as global context cues. By collecting the condensed contexts, we employ a Transformer decoder to investigate the relations between them and each local feature from the FP and then refine the MFF results accordingly. As a result, we obtain a simple and light-weight Transformer-based Context Condensation (TCC) module, which can boost various FPs and lower their computational costs simultaneously. Extensive experimental results on the challenging MS COCO dataset show that TCC is compatible to four representative FPs and consistently improves their detection accuracy by up to 7.8 % in terms of average precision and reduce their complexities by up to around 20% in terms of GFLOPs, helping them achieve state-of-the-art performance more efficiently. Code will be released.

cs.CV↗

arXiv2022

Coloured gravitational instantons, the strong CP problem and the companion axion solution

Quantum gravity introduces a new source of the combined parity (CP) violation in gauge theories. We argue that this new CP violation gets bundled with the strong CP violation through the coloured gravitational instantons. Consequently, the standard axion solution to the strong CP problem is compromised. Further, we argue that the ultimate solution to the strong CP problem must involve at least one additional axion particle.

hep-ph↗

arXiv2022

Efficient Attribute-Based Smart Contract Access Control Enhanced by Reputation Assessment

Blockchain's immutability can resist unauthorized changes of ledgers, thus it can be used as a trust enhancement mechanism to a shared system. Indeed, blockchain has been considered to solve the security and privacy issues of the Internet of Things (IoT). In this regard, most researches currently focus on the realization of various access control models and architectures, and are working towards making full use of the blockchain to secure IoT systems. It is worth noting that there has been an increasingly heavy pressure on the blockchain storage caused by dealing with massive IoT data and handling malicious access behaviors in the system, and not many countermeasures have been seen to curb the increase. However, this problem has not been paid enough attention. In this paper, we implement an attribute-based access control scheme using smart contracts in Quorum blockchain. It provides basic access control functions and conserves storage by reducing the number of smart contracts. In addition, a reputation-based technique is introduced to cope with malicious behaviors. Certain illegal transactions can be blocked by the credit-assessment algorithm, which deters possibly malicious nodes and gives more chance to well-behaved nodes. The feasibility of our proposed scheme is demonstrated by doing experiment on a testbed and conducting a case study. Finally, the system performance is assessed based on experimental measurement.

cs.CR↗

arXiv2022

Dropout Prediction Uncertainty Estimation Using Neuron Activation Strength

Dropout has been commonly used to quantify prediction uncertainty, i.e, the variations of model predictions on a given input example. However, using dropout in practice can be expensive as it requires running dropout inferences many times. In this paper, we study how to estimate dropout prediction uncertainty in a resource-efficient manner. We demonstrate that we can use neuron activation strengths to estimate dropout prediction uncertainty under different dropout settings and on a variety of tasks using three large datasets, MovieLens, Criteo, and EMNIST. Our approach provides an inference-once method to estimate dropout prediction uncertainty as a cheap auxiliary task. We also demonstrate that using activation features from a subset of the neural network layers can be sufficient to achieve uncertainty estimation performance almost comparable to that of using activation features from all layers, thus reducing resources even further for uncertainty estimation.

cs.LG↗