Search Results for author: Cheng-Zhong Xu

Found 44 papers, 19 papers with code

DME-Driver: Integrating Human Decision Logic and 3D Scene Perception in Autonomous Driving

no code implementations8 Jan 2024 Wencheng Han, Dongqian Guo, Cheng-Zhong Xu, Jianbing Shen

On the other hand, the generation of accurate control signals relies on precise and detailed environmental perception, which is where 3D scene perception models excel.

Autonomous Driving Language Modelling +1

Leveraging Frame Affinity for sRGB-to-RAW Video De-rendering

no code implementations CVPR 2024 Chen Zhang, Wencheng Han, Yang Zhou, Jianbing Shen, Cheng-Zhong Xu, Wentao Liu

These methods utilize both the metadata and the sRGB image to perform sRGB-to-RAW de-rendering and recover high-quality single-frame RAW data.

Image Reconstruction Video Editing +1

Truth Forest: Toward Multi-Scale Truthfulness in Large Language Models through Intervention without Tuning

1 code implementation29 Dec 2023 Zhongzhi Chen, Xingwu Sun, Xianfeng Jiao, Fengzong Lian, Zhanhui Kang, Di Wang, Cheng-Zhong Xu

We introduce Truth Forest, a method that enhances truthfulness in LLMs by uncovering hidden truth representations using multi-dimensional orthogonal probes.

DI-V2X: Learning Domain-Invariant Representation for Vehicle-Infrastructure Collaborative 3D Object Detection

1 code implementation25 Dec 2023 Li Xiang, Junbo Yin, Wei Li, Cheng-Zhong Xu, Ruigang Yang, Jianbing Shen

Specifically, DMA builds a domain-mixing 3D instance bank for the teacher and student models during training, resulting in aligned data representation.

3D Object Detection object-detection +1

APBench: A Unified Benchmark for Availability Poisoning Attacks and Defenses

1 code implementation7 Aug 2023 Tianrui Qin, Xitong Gao, Juanjuan Zhao, Kejiang Ye, Cheng-Zhong Xu

To further evaluate the attack and defense capabilities of these poisoning methods, we have developed a benchmark -- APBench for assessing the efficacy of adversarial poisoning.

Data Augmentation

GSB: Group Superposition Binarization for Vision Transformer with Limited Training Samples

1 code implementation13 May 2023 Tian Gao, Cheng-Zhong Xu, Le Zhang, Hui Kong

Compared with the full-precision one, the model with the binarization method replaces complex tensor multiplication with simple bit-wise binary operations and represents full-precision model parameters and activations with only 1-bit ones, which potentially solves the problem of model size and computational complexity, respectively.

Binarization Knowledge Distillation +2

Learning the Unlearnable: Adversarial Augmentations Suppress Unlearnable Example Attacks

1 code implementation27 Mar 2023 Tianrui Qin, Xitong Gao, Juanjuan Zhao, Kejiang Ye, Cheng-Zhong Xu

In this paper, we introduce the UEraser method, which outperforms current defenses against different types of state-of-the-art unlearnable example attacks through a combination of effective data augmentation policies and loss-maximizing adversarial augmentations.

Data Augmentation Data Poisoning

LiDAR-CS Dataset: LiDAR Point Cloud Dataset with Cross-Sensors for 3D Object Detection

1 code implementation29 Jan 2023 Jin Fang, Dingfu Zhou, Jingjing Zhao, Chenming Wu, Chulin Tang, Cheng-Zhong Xu, Liangjun Zhang

This setting results in two distinct domain gaps: scenarios and sensors, making it difficult to analyze and evaluate the method accurately.

3D Object Detection Autonomous Driving +2

Efficient Loss Function by Minimizing the Detrimental Effect of Floating-Point Errors on Gradient-Based Attacks

no code implementations CVPR 2023 Yunrui Yu, Cheng-Zhong Xu

Attackers can deceive neural networks by adding human imperceptive perturbations to their input data; this reveals the vulnerability and weak robustness of current deep-learning networks.

Flareon: Stealthy any2any Backdoor Injection via Poisoned Augmentation

1 code implementation20 Dec 2022 Tianrui Qin, Xianghuan He, Xitong Gao, Yiren Zhao, Kejiang Ye, Cheng-Zhong Xu

Open software supply chain attacks, once successful, can exact heavy costs in mission-critical applications.

Data Augmentation

MORA: Improving Ensemble Robustness Evaluation with Model-Reweighing Attack

1 code implementation15 Nov 2022 Yunrui Yu, Xitong Gao, Cheng-Zhong Xu

In particular, most ensemble defenses exhibit near or exactly 0% robustness against MORA with $\ell^\infty$ perturbation within 0. 02 on CIFAR-10, and 0. 01 on CIFAR-100.

Adversarial Attack

Semi-supervised 3D Object Detection with Proficient Teachers

1 code implementation26 Jul 2022 Junbo Yin, Jin Fang, Dingfu Zhou, Liangjun Zhang, Cheng-Zhong Xu, Jianbing Shen, Wenguan Wang

To reduce the dependence on large supervision, semi-supervised learning (SSL) based approaches have been proposed.

3D Object Detection Autonomous Driving +3

ProposalContrast: Unsupervised Pre-training for LiDAR-based 3D Object Detection

1 code implementation26 Jul 2022 Junbo Yin, Dingfu Zhou, Liangjun Zhang, Jin Fang, Cheng-Zhong Xu, Jianbing Shen, Wenguan Wang

Existing approaches for unsupervised point cloud pre-training are constrained to either scene-level or point/voxel-level instance discrimination.

3D Object Detection object-detection +2

Improving Pre-trained Language Model Fine-tuning with Noise Stability Regularization

no code implementations12 Jun 2022 Hang Hua, Xingjian Li, Dejing Dou, Cheng-Zhong Xu, Jiebo Luo

The advent of large-scale pre-trained language models has contributed greatly to the recent progress in natural language processing.

Domain Generalization Language Modelling +3

Unsupervised Visible-light Images Guided Cross-Spectrum Depth Estimation from Dual-Modality Cameras

no code implementations30 Apr 2022 Yubin Guo, Haobo Jiang, Xinlei Qi, Jin Xie, Cheng-Zhong Xu, Hui Kong

Meanwhile, we release a large dual-spectrum depth estimation dataset with visible-light and far-infrared stereo images captured in different scenes to the society.

Depth Estimation

FedDC: Federated Learning with Non-IID Data via Local Drift Decoupling and Correction

1 code implementation CVPR 2022 Liang Gao, Huazhu Fu, Li Li, YingWen Chen, Ming Xu, Cheng-Zhong Xu

Federated learning (FL) allows multiple clients to collectively train a high-performance global model without sharing their private data.

Federated Learning Image Classification

Grounding Commands for Autonomous Vehicles via Layer Fusion with Region-specific Dynamic Layer Attention

no code implementations14 Mar 2022 Hou Pong Chan, Mingxi Guo, Cheng-Zhong Xu

In this work, we study the problem of language grounding for autonomous vehicles, which aims to localize a region in a visual scene according to a natural language command from a passenger.

Autonomous Vehicles

Boosting Active Learning via Improving Test Performance

1 code implementation10 Dec 2021 Tianyang Wang, Xingjian Li, Pengkun Yang, Guosheng Hu, Xiangrui Zeng, Siyu Huang, Cheng-Zhong Xu, Min Xu

In this work, we explore such an impact by theoretically proving that selecting unlabeled data of higher gradient norm leads to a lower upper-bound of test loss, resulting in better test performance.

Active Learning Electron Tomography +2

SenseMag: Enabling Low-Cost Traffic Monitoring using Non-invasive Magnetic Sensing

no code implementations24 Oct 2021 Kafeng Wang, Haoyi Xiong, Jie Zhang, Hongyang Chen, Dejing Dou, Cheng-Zhong Xu

Extensive experiment based on real-word field deployment (on the highways in Shenzhen, China) shows that SenseMag significantly outperforms the existing methods in both classification accuracy and the granularity of vehicle types (i. e., 7 types by SenseMag versus 4 types by the existing work in comparisons).


FedDrop: Trajectory-weighted Dropout for Efficient Federated Learning

no code implementations29 Sep 2021 Dongping Liao, Xitong Gao, Yiren Zhao, Hao Dai, Li Li, Kafeng Wang, Kejiang Ye, Yang Wang, Cheng-Zhong Xu

Federated learning (FL) enables edge clients to train collaboratively while preserving individual's data privacy.

Federated Learning

LAFEAT: Piercing Through Adversarial Defenses with Latent Features

1 code implementation CVPR 2021 Yunrui Yu, Xitong Gao, Cheng-Zhong Xu

In this paper, we show that latent features in certain "robust" models are surprisingly susceptible to adversarial attacks.

Adaptive Consistency Regularization for Semi-Supervised Transfer Learning

1 code implementation CVPR 2021 Abulikemu Abuduweili, Xingjian Li, Humphrey Shi, Cheng-Zhong Xu, Dejing Dou

To better exploit the value of both pre-trained weights and unlabeled target examples, we introduce adaptive consistency regularization that consists of two complementary components: Adaptive Knowledge Consistency (AKC) on the examples between the source and target model, and Adaptive Representation Consistency (ARC) on the target model between labeled and unlabeled examples.

Pseudo Label Transfer Learning

Frontier Detection and Reachability Analysis for Efficient 2D Graph-SLAM Based Active Exploration

1 code implementation7 Sep 2020 Zezhou Sun, Banghe Wu, Cheng-Zhong Xu, Sanjay E. Sarma, Jian Yang, Hui Kong

We propose an integrated approach to active exploration by exploiting the Cartographer method as the base SLAM module for submap creation and performing efficient frontier detection in the geometrically co-aligned submaps induced by graph optimization.

XMixup: Efficient Transfer Learning with Auxiliary Samples by Cross-domain Mixup

no code implementations20 Jul 2020 Xingjian Li, Haoyi Xiong, Haozhe An, Cheng-Zhong Xu, Dejing Dou

While the existing multitask learning algorithms need to run backpropagation over both the source and target datasets and usually consume a higher gradient complexity, XMixup transfers the knowledge from source to target tasks more efficiently: for every class of the target task, XMixup selects the auxiliary samples from the source dataset and augments training samples via the simple mixup strategy.

Transfer Learning

RIFLE: Backpropagation in Depth for Deep Transfer Learning through Re-Initializing the Fully-connected LayEr

1 code implementation ICML 2020 Xingjian Li, Haoyi Xiong, Haozhe An, Cheng-Zhong Xu, Dejing Dou

RIFLE brings meaningful updates to the weights of deep CNN layers and improves low-level feature learning, while the effects of randomization can be easily converged throughout the overall learning procedure.

Transfer Learning

How to Build a Graph-Based Deep Learning Architecture in Traffic Domain: A Survey

no code implementations24 May 2020 Jiexia Ye, Juanjuan Zhao, Kejiang Ye, Cheng-Zhong Xu

Recently, various novel deep learning techniques have been developed to process graph data, called graph neural networks (GNNs).

Multi-Graph Convolutional Network for Relationship-Driven Stock Movement Prediction

1 code implementation11 May 2020 Jiexia Ye, Juanjuan Zhao, Kejiang Ye, Cheng-Zhong Xu

However, it is well known that an individual stock price is correlated with prices of other stocks in complex ways.

Stock Prediction

Pay Attention to Features, Transfer Learn Faster CNNs

no code implementations ICLR 2020 Kafeng Wang, Xitong Gao, Yiren Zhao, Xingjian Li, Dejing Dou, Cheng-Zhong Xu

Deep convolutional neural networks are now widely deployed in vision applications, but a limited size of training data can restrict their task performance.

Transfer Learning

LiDAR Iris for Loop-Closure Detection

no code implementations9 Dec 2019 Ying Wang, Zezhou Sun, Cheng-Zhong Xu, Sanjay Sarma, Jian Yang, Hui Kong

In this paper, a global descriptor for a LiDAR point cloud, called LiDAR Iris, is proposed for fast and accurate loop-closure detection.

Loop Closure Detection

A Robust Stereo Camera Localization Method with Prior LiDAR Map Constrains

no code implementations2 Dec 2019 Dong Han, Zuhao Zou, Lujia Wang, Cheng-Zhong Xu

Different from the conventional visual localization system, we design a novel visual optimization model by matching planar information between the LiDAR map and visual image.

Camera Localization Visual Localization

Focused Quantization for Sparse CNNs

1 code implementation NeurIPS 2019 Yiren Zhao, Xitong Gao, Daniel Bates, Robert Mullins, Cheng-Zhong Xu

In ResNet-50, we achieved a 18. 08x CR with only 0. 24% loss in top-5 accuracy, outperforming existing compression methods.

Neural Network Compression Quantization

Dynamic Channel Pruning: Feature Boosting and Suppression

2 code implementations ICLR 2019 Xitong Gao, Yiren Zhao, Łukasz Dudziak, Robert Mullins, Cheng-Zhong Xu

Making deep convolutional neural networks more accurate typically comes at the cost of increased computational and memory resources.

Model Compression Network Pruning

Cannot find the paper you are looking for? You can Submit a new open access paper.