no code implementations • Findings (EMNLP) 2021 • Kezhen Chen, Qiuyuan Huang, Daniel McDuff, Xiang Gao, Hamid Palangi, JianFeng Wang, Kenneth Forbus, Jianfeng Gao
Based on these annotations, we define two different tasks for the NICE dataset.
no code implementations • 24 Apr 2024 • Xiang Gao, Yuqi Zhang
The saliency map is utilized for content regularization from two aspects, both explicitly and implicitly: (\romannumeral1) we propose saliency IOU (SIOU) loss to explicitly regularize saliency consistency before and after stylization; (\romannumeral2) we propose saliency adaptive normalization (SANorm) which implicitly enhances content integrity of the generated paintings by injecting saliency information to the generator network to guide painting generation.
Generative Adversarial Network Image-to-Image Translation +3
no code implementations • 4 Mar 2024 • Xiang Gao, Jiaxin Zhang, Lalla Mouatadid, Kamalika Das
Motivated by this gap, we introduce a novel UQ method, sampling with perturbation for UQ (SPUQ), designed to tackle both aleatoric and epistemic uncertainties.
1 code implementation • 27 Feb 2024 • Bi'an Du, Xiang Gao, Wei Hu, Renjie Liao
Subsequently, we transform the point cloud using the latent poses, feeding it to the part encoder for aggregating super-part information and reasoning about part relationships to predict all part poses.
1 code implementation • 8 Feb 2024 • Mingyi Zhou, Xiang Gao, Jing Wu, Kui Liu, Hailong Sun, Li Li
Our findings emphasize the need for developers to carefully consider their model deployment strategies, and use white-box methods to evaluate the vulnerability of on-device models.
no code implementations • 30 Jan 2024 • Xiang Gao, Kamalika Das
However, it can be challenging to align LLMs with our intent, particularly when we want to generate content that is preferable over others or when we want the LLM to respond in a certain style or tone that is hard to describe.
no code implementations • 8 Jan 2024 • Zhimin Zhang, Xiang Gao, Wei Hu
The convenience of 3D sensors has led to an increase in the use of 3D point clouds in various applications.
no code implementations • 5 Dec 2023 • Sandeep Chinta, Xiang Gao, Qing Zhu
One of the main sources of this uncertainty arises from the numerous uncertain model parameters within various physical, biological, and chemical processes that influence methane production, oxidation, and transport.
no code implementations • 17 Oct 2023 • Ziqing Zhu, Xiang Gao, Siqi Bu, Ka Wing Chan, Bin Zhou, Shiwei Xia
This online algorithm enables MGs to spontaneously search for the Pareto Frontier considering multiple objectives and risk mitigation.
1 code implementation • 9 Oct 2023 • Chunge Bai, Ruijie Fu, Xiang Gao
In contrast, mapping methods based on LiDAR scans are popular in large-scale urban scene reconstruction due to their precise distance measurements, a capability fundamentally absent in visual-based approaches.
no code implementations • 29 Sep 2023 • Xiang Gao, Hainan Cui, Shuhan Shen
In addition, to further address the limitations of the existing rotation averaging benchmark of relying on the slightly outdated Bundler camera calibration results as ground truths and focusing solely on rotation estimation accuracy, this paper presents a new COLMAP-based rotation averaging benchmark that incorporates a cross check between COLMAP and Bundler, and employ the accuracy of both rotation and downstream location estimation as evaluation metrics, which is desired to provide a more reliable and comprehensive evaluation tool for the rotation averaging research.
no code implementations • 18 Jul 2023 • Jianzhu Huai, Xiang Gao
In contrast, the iterated EKF (IEKF) refines the state in the update step by iteratively solving a least squares problem.
1 code implementation • 15 Jun 2023 • Binhang Qi, Hailong Sun, Hongyu Zhang, Ruobing Zhao, Xiang Gao
In this paper, we propose a novel approach that incorporates modularization into the model training process, i. e., modularizing-while-training (MwT).
1 code implementation • 8 Jun 2023 • Jun Zhao, Xin Zhao, WenYu Zhan, Qi Zhang, Tao Gui, Zhongyu Wei, Yunwen Chen, Xiang Gao, Xuanjing Huang
Inspired by text adversarial attacks, we adaptively apply small but critical perturbations to original training instances and thus synthesizing negative instances that are more likely to be mistaken by the model as known relations.
1 code implementation • 22 May 2023 • Xiao Wang, Weikang Zhou, Qi Zhang, Jie zhou, Songyang Gao, Junzhe Wang, Menghan Zhang, Xiang Gao, Yunwen Chen, Tao Gui
Pretrained language models have achieved remarkable success in various natural language processing tasks.
1 code implementation • 1 Apr 2023 • Binhang Qi, Hailong Sun, Xiang Gao, Hongyu Zhang, Zhaotian Li, Xudong Liu
Prior approaches to DNN model reuse have two main limitations: 1) reusing the entire model, while only a small part of the model's functionalities (labels) are required, would cause much overhead (e. g., computational and time costs for inference), and 2) model reuse would inherit the defects and weaknesses of the reused model, and hence put the new system under threats of security attack.
no code implementations • 23 Nov 2022 • Xiang Gao, Weihao Gao, Wenzhi Xiao, Zhirui Wang, Chong Wang, Liang Xiang
Experiments show that, compared to training from scratch, fine-tuning the pretrained model can significantly improve the performance for seven molecular property prediction tasks and two force field tasks.
no code implementations • 23 Nov 2022 • Xiang Gao, Weihao Gao, Wenzhi Xiao, Zhirui Wang, Chong Wang, Liang Xiang
To model the complex nonlinearity in predicting molecular properties in an more end-to-end approach, we propose to encode the positional quantities with a learnable embedding that is continuous and differentiable.
no code implementations • 14 Nov 2022 • Xiang Gao, Wei Hu, Renjie Liao
The decoder takes the latent variable and the feature from the encoder as an input and predicts the per-point part distribution at the top level.
2 code implementations • Proceedings of the 30th ACM International Conference on Multimedia 2022 • Wenke Huang, Mang Ye, Bo Du, Xiang Gao
To address these issues, this paper presents a novel framework with two main parts: 1) model agnostic federated learning, it performs public-private communication by unifying the model prediction outputs on the shared public datasets; 2) latent embedding adaptation, it addresses the domain gap with an adversarial learning scheme to discriminate the public and private domains.
no code implementations • 10 Oct 2022 • Ruotian Ma, Xuanting Chen, Lin Zhang, Xin Zhou, Junzhe Wang, Tao Gui, Qi Zhang, Xiang Gao, Yunwen Chen
In this work, we conduct an empirical study on the "Unlabeled Entity Problem" and find that it leads to severe confusion between "O" and entities, decreasing class discrimination of old classes and declining the model's ability to learn new classes.
1 code implementation • 11 Sep 2022 • Binhang Qi, Hailong Sun, Xiang Gao, Hongyu Zhang
To patch a weak CNN model that performs unsatisfactorily on a target class (TC), we compose the weak CNN model with the corresponding module obtained from a strong CNN model.
1 code implementation • 2 Aug 2022 • Xiang Gao, Yuqi Zhang, Yingjie Tian
Image cartoonization is recently dominated by generative adversarial networks (GANs) from the perspective of unsupervised image-to-image translation, in which an inherent challenge is to precisely capture and sufficiently transfer characteristic cartoon styles (e. g., clear edges, smooth color shading, abstract fine structures, etc.).
no code implementations • 26 Jun 2022 • Xiang Gao, Yingjie Tian, Zhiquan Qi
We propose an end-to-end-trainable feature augmentation module built for image classification that extracts and exploits multi-view local features to boost model performance.
no code implementations • 9 Feb 2022 • Xiang Gao, Cody Hyndman, Traian A. Pirvu, Petar Jevtić
In this paper, we study the problem of post-retirement annuitization with extra labor income in the framework of stochastic control, optimal stopping, and expected utility maximization.
no code implementations • 8 Jan 2022 • Qi Qi, Kunqian Li, Haiyong Zheng, Xiang Gao, Guojia Hou, Kun Sun
In this paper, we propose a novel underwater image enhancement network, called SGUIE-Net, in which we introduce semantic information as high-level guidance across different images that share common semantic regions.
no code implementations • 5 Aug 2021 • Wei-Wen Hsu, Yongfang Wu, Chang Hao, Yu-Ling Hou, Xiang Gao, Yun Shao, Xueli Zhang, Tao He, Yanhong Tai
Objective: We develop a computer-aided diagnosis (CAD) system using deep learning approaches for lesion detection and classification on whole-slide images (WSIs) with breast cancer.
no code implementations • 5 Jul 2021 • Bi'an Du, Xiang Gao, Wei Hu, Xin Li
Point clouds have attracted increasing attention.
1 code implementation • 25 May 2021 • Xiang Gao, Wei Hu, Guo-Jun Qi
We formalize the proposed model from an information-theoretic perspective, by maximizing the mutual information between topology transformations and node representations before and after the transformations.
1 code implementation • 14 May 2021 • Yizhe Zhang, Siqi Sun, Xiang Gao, Yuwei Fang, Chris Brockett, Michel Galley, Jianfeng Gao, Bill Dolan
We propose a framework that alleviates this data constraint by jointly training a grounded generator and document retriever on the language model signal.
1 code implementation • 16 Apr 2021 • Xiang Gao, Yizhe Zhang, Michel Galley, Bill Dolan
To alleviate this risk, we propose an adversarial training approach to learn a robust model, ATT (Adversarial Turing Test), that discriminates machine-generated responses from human-written replies.
no code implementations • 1 Mar 2021 • Xiang Gao, Wei Hu, Guo-Jun Qi
Then, we self-train a representation to capture the intrinsic 3D object representation by decoding 3D transformation parameters from the fused feature representations of multiple views before and after the transformation.
no code implementations • 11 Feb 2021 • Xiang Gao, Nikhil Karthik, Swagato Mukherjee, Peter Petreczky, Sergey Syritsyn, Yong Zhao
We study the form factor at the physical point with a lattice spacing $a=0. 076$ fm.
High Energy Physics - Lattice High Energy Physics - Experiment High Energy Physics - Phenomenology Nuclear Theory
no code implementations • 27 Jan 2021 • Xiang Gao, Nikhil Karthik, Swagato Mukherjee, Peter Petreczky, Sergey Syritsyn, Yong Zhao
We present an exploratory lattice QCD investigation of the differences between the valence quark structure of pion and its radial excitation $\pi(1300)$ in a fixed finite volume using the leading-twist factorization approach.
High Energy Physics - Lattice High Energy Physics - Experiment High Energy Physics - Phenomenology Nuclear Theory
no code implementations • 1 Jan 2021 • Xiang Gao, Wei Hu, Guo-Jun Qi
We formalize the TopoTER from an information-theoretic perspective, by maximizing the mutual information between topology transformations and node representations before and after the transformations.
2 code implementations • EMNLP 2020 • Xiang Gao, Yizhe Zhang, Michel Galley, Chris Brockett, Bill Dolan
Particularly, our ranker outperforms the conventional dialog perplexity baseline with a large margin on predicting Reddit feedback.
1 code implementation • ACL 2020 • Yizhe Zhang, Siqi Sun, Michel Galley, Yen-Chun Chen, Chris Brockett, Xiang Gao, Jianfeng Gao, Jingjing Liu, Bill Dolan
We present a large, tunable neural conversational response generation model, DIALOGPT (dialogue generative pre-trained transformer).
no code implementations • 16 Jun 2020 • Xiang Gao, Jennie Si, Yue Wen, Minhan Li, He, Huang
We are motivated by the real challenges presented in a human-robot system to develop new designs that are efficient at data level and with performance guarantees such as stability and optimality at systems level.
no code implementations • 11 Jun 2020 • Minhan Li, Yue Wen, Xiang Gao, Jennie Si, He Helen Huang
Personalizing medical devices such as lower limb wearable robots is challenging.
1 code implementation • ACL 2020 • Xiang Gao, Michel Galley, Bill Dolan
We present MixingBoard, a platform for quickly building demos with a focus on knowledge grounded stylized text generation.
1 code implementation • 1 May 2020 • Zeqiu Wu, Michel Galley, Chris Brockett, Yizhe Zhang, Xiang Gao, Chris Quirk, Rik Koncel-Kedziorski, Jianfeng Gao, Hannaneh Hajishirzi, Mari Ostendorf, Bill Dolan
Current end-to-end neural conversation models inherently lack the flexibility to impose semantic control in the response generation process, often resulting in uninteresting responses.
1 code implementation • EMNLP 2020 • Chunyuan Li, Xiang Gao, Yuan Li, Baolin Peng, Xiujun Li, Yizhe Zhang, Jianfeng Gao
We hope that our first pre-trained big VAE language model itself and results can help the NLP community renew the interests of deep generative models in the era of large-scale pre-training, and make these principled methods more practical.
no code implementations • 17 Mar 2020 • Wei Hu, Qianjiang Hu, Zehua Wang, Xiang Gao
In particular, we define a manifold-to-manifold distance and its discrete counterpart on graphs to measure the variation-based intrinsic distance between surface patches in the temporal domain, provided that graph operators are discrete counterparts of functionals on Riemannian manifolds.
1 code implementation • CVPR 2020 • Xiang Gao, Wei Hu, Guo-Jun Qi
Recent advances in Graph Convolutional Neural Networks (GCNNs) have shown their efficiency for non-Euclidean data on graphs, which often require a large amount of labeled data with high cost.
6 code implementations • 1 Nov 2019 • Yizhe Zhang, Siqi Sun, Michel Galley, Yen-Chun Chen, Chris Brockett, Xiang Gao, Jianfeng Gao, Jingjing Liu, Bill Dolan
We present a large, tunable neural conversational response generation model, DialoGPT (dialogue generative pre-trained transformer).
1 code implementation • 11 Sep 2019 • Jiaxiang Tang, Wei Hu, Xiang Gao, Zongming Guo
In particular, we cast the graph optimization problem as distance metric learning to capture pairwise similarities of features in each layer.
1 code implementation • IJCNLP 2019 • Xiang Gao, Yizhe Zhang, Sungjin Lee, Michel Galley, Chris Brockett, Jianfeng Gao, Bill Dolan
This structure allows the system to generate stylized relevant responses by sampling in the neighborhood of the conversation model prediction, and continuously control the style level.
no code implementations • 22 Jul 2019 • Wei Hu, Xiang Gao, Gene Cheung, Zongming Guo
In this work, we assume instead the availability of a relevant feature vector $\mathbf{f}_i$ per node $i$, from which we compute an optimal feature graph via optimization of a feature metric.
no code implementations • ACL 2019 • Vighnesh Leonardo Shiv, Chris Quirk, Anshuman Suri, Xiang Gao, Khuram Shahid, Nithya Govindarajan, Yizhe Zhang, Jianfeng Gao, Michel Galley, Chris Brockett, Tulasi Menon, Bill Dolan
The Intelligent Conversation Engine: Code and Pre-trained Systems (Microsoft Icecaps) is an upcoming open-source natural language processing repository.
1 code implementation • ACL 2019 • Lianhui Qin, Michel Galley, Chris Brockett, Xiaodong Liu, Xiang Gao, Bill Dolan, Yejin Choi, Jianfeng Gao
Although neural conversation models are effective in learning how to produce fluent responses, their primary challenge lies in knowing what to say to make the conversation contentful and non-vacuous.
no code implementations • 28 Apr 2019 • Wei Hu, Qianjiang Hu, Zehua Wang, Xiang Gao
Finally, based on the spatial-temporal graph learning, we formulate dynamic point cloud denoising as the joint optimization of the desired point cloud and underlying spatio-temporal graph, which leverages both intra-frame affinities and inter-frame consistency and is solved via alternating minimization.
no code implementations • 23 Apr 2019 • Xiang Gao, Wei Hu, Zongming Guo
In this paper, we propose Graph Learning Neural Networks (GLNNs), which exploit the optimization of graphs (the adjacency matrix in particular) from both data and tasks.
no code implementations • 21 Apr 2019 • Xiang Gao, Shuhan Shen, Lingjie Zhu, Tianxin Shi, Zhiheng Wang, Zhanyi Hu
Experimental evaluations on two ancient Chinese architecture datasets demonstrate the effectiveness of our proposed complete scene reconstruction pipeline.
no code implementations • 17 Apr 2019 • Jiabin Xue, Jiqing Han, Tieran Zheng, Xiang Gao, Jiaxing Guo
On the one hand, we constrain the new parameters not to deviate too far from the original parameters and punish the new system when forgetting original knowledge.
Automatic Speech Recognition Automatic Speech Recognition (ASR) +2
no code implementations • 8 Apr 2019 • Tianxin Shi, Shuhan Shen, Xiang Gao, Lingjie Zhu
Accurate and robust visual localization under a wide range of viewing condition variations including season and illumination changes, as well as weather and day-night variations, is the key component for many computer vision and robotics applications.
1 code implementation • 13 Mar 2019 • Yizhe Zhang, Xiang Gao, Sungjin Lee, Chris Brockett, Michel Galley, Jianfeng Gao, Bill Dolan
Generating responses that are consistent with the dialogue context is one of the central challenges in building engaging conversational agents.
no code implementations • 4 Mar 2019 • Wei-Wen Hsu, Chung-Hao Chen, Chang Hoa, Yu-Ling Hou, Xiang Gao, Yun Shao, Xueli Zhang, Jingjing Wang, Tao He, Yanghong Tai
Most of the characteristics learned by the deep learning models have summarized the detection rules that can be recognized by the experienced pathologists, whereas there are still some features may not be intuitive to domain experts but discriminative in classification for machines.
no code implementations • NAACL 2019 • Xiang Gao, Sungjin Lee, Yizhe Zhang, Chris Brockett, Michel Galley, Jianfeng Gao, Bill Dolan
In this paper, we propose a SpaceFusion model to jointly optimize diversity and relevance that essentially fuses the latent space of a sequence-to-sequence model and that of an autoencoder model by leveraging novel regularization terms.
Ranked #1 on Dialogue Generation on Reddit (multi-ref)
no code implementations • 11 Jan 2019 • Koichiro Yoshino, Chiori Hori, Julien Perez, Luis Fernando D'Haro, Lazaros Polymenakos, Chulaka Gunasekara, Walter S. Lasecki, Jonathan K. Kummerfeld, Michel Galley, Chris Brockett, Jianfeng Gao, Bill Dolan, Xiang Gao, Huda Alamari, Tim K. Marks, Devi Parikh, Dhruv Batra
This paper introduces the Seventh Dialog System Technology Challenges (DSTC), which use shared datasets to explore the problem of building dialog systems.
no code implementations • 29 Nov 2018 • Xiang Gao, Wei Hu, Jiaxiang Tang, Jiaying Liu, Zongming Guo
In this paper, we represent skeletons naturally on graphs, and propose a graph regression based GCN (GR-GCN) for skeleton-based action recognition, aiming to capture the spatio-temporal variation in the data.
Ranked #2 on Skeleton Based Action Recognition on Florence 3D
no code implementations • ECCV 2018 • Lingjie Zhu, Shuhan Shen, Xiang Gao, Zhanyi Hu
There are two major steps in our framework: segmentation and building modeling.
no code implementations • 3 Aug 2018 • Xiang Gao, Rui Wang, Nikolaus Demmel, Daniel Cremers
In this paper we present an extension of Direct Sparse Odometry (DSO) to a monocular visual SLAM system with loop closure detection and pose-graph optimization (LDSO).
no code implementations • 1 Aug 2018 • Sheng Chen, Jia Guo, Yang Liu, Xiang Gao, Zhen Han
In this paper, we propose a novel Global Norm-Aware Pooling (GNAP) block, which reweights local features in a convolutional neural network (CNN) adaptively according to their L2 norms and outputs a global feature vector with a global average pooling layer.
17 code implementations • 20 Apr 2018 • Sheng Chen, Yang Liu, Xiang Gao, Zhen Han
Face Analysis Project on MXNet
Ranked #5 on Lightweight Face Recognition on AgeDB-30
no code implementations • 23 Mar 2018 • Hainan Cui, Shuhan Shen, Xiang Gao, Zhanyi Hu
The global manner has the advantage of simultaneously estimating all camera poses, but it is usually sensitive to epipolar geometry outliers.
2 code implementations • 11 Mar 2018 • Xiang Gao
Deep Q-learning is investigated as an end-to-end solution to estimate the optimal strategies for acting on time series input.
no code implementations • CVPR 2017 • Hainan Cui, Xiang Gao, Shuhan Shen, Zhanyi Hu
In this work, we propose a new hybrid SfM method to tackle the issues of efficiency, accuracy and robustness in a unified framework.
no code implementations • 11 May 2017 • Nan Yang, Rui Wang, Xiang Gao, Daniel Cremers
Monocular visual odometry (VO) and simultaneous localization and mapping (SLAM) have seen tremendous improvements in accuracy, robustness and efficiency, and have gained increasing popularity over recent years.
Monocular Visual Odometry Simultaneous Localization and Mapping
no code implementations • 19 May 2016 • Xiang Gao, Yangyang Xu, Shuzhong Zhang
Assuming mere convexity, we establish its $O(1/t)$ convergence rate in terms of the objective value and feasibility measure.