1 code implementation • 23 Nov 2024 • Jilong Guo, Haobo Yang, Mo Zhou, Xinyu Zhang
This method constructs specific masks based on gradient fluctuations to isolate parameters influenced by other tasks, ensuring that the model achieves strong performance across all scenarios without adding extra parameters.
no code implementations • 17 Nov 2024 • Lei Yang, Xinyu Zhang, Jun Li, Chen Wang, Zhiying Song, Tong Zhao, Ziying Song, Li Wang, Mo Zhou, Yang shen, Kai Wu, Chen Lv
Previous studies have demonstrated the effectiveness of cooperative perception in extending the perception range and overcoming occlusions, thereby improving the safety of autonomous driving.
no code implementations • 24 Sep 2024 • Mo Zhou, Stanley Osher, Wuchen Li
Classical neural ordinary differential equations (ODEs) are powerful tools for approximating the log-density functions in high-dimensional spaces along trajectories, where neural networks parameterize the velocity fields.
no code implementations • 3 Jun 2024 • Mo Zhou, Rong Ge
The ability of learning useful features is one of the major advantages of neural networks.
no code implementations • 27 Feb 2024 • Mo Zhou, Yiding Yang, Haoxiang Li, Vishal M. Patel, Gang Hua
With a strong alignment between the training and test distributions, object relation as a context prior facilitates object detection.
1 code implementation • 3 Feb 2024 • Yatong Bai, Mo Zhou, Vishal M. Patel, Somayeh Sojoudi
Adversarial robustness often comes at the cost of degraded accuracy, impeding real-life applications of robust classification models.
1 code implementation • 25 May 2023 • Yu Zeng, Mo Zhou, Yuan Xue, Vishal M. Patel
Prior research attempted to mitigate these threats by detecting generated images, but the varying traces left by different generative models make it challenging to create a universal detector capable of generalizing to new, unseen generative models.
no code implementations • 24 May 2023 • Kangfu Mei, Mo Zhou, Vishal M. Patel
The model can be scaled to generate high-resolution data while unifying multiple modalities.
no code implementations • 3 Apr 2023 • Yunwei Ren, Mo Zhou, Rong Ge
Depth separation -- why a deeper network is more powerful than a shallower one -- has been a major problem in deep learning theory.
no code implementations • 11 Feb 2023 • Mo Zhou, Jianfeng Lu
We consider policy gradient methods for stochastic optimal control problem in continuous time.
no code implementations • 1 Feb 2023 • Mo Zhou, Rong Ge
In this work, we give a different parametrization of the model which leads to a new implicit regularization effect that combines the benefit of $\ell_1$ and $\ell_2$ interpolators.
1 code implementation • 16 Dec 2022 • Mo Zhou, Jiequn Han, Manas Rachh, Carlos Borges
We present a neural network warm-start approach for solving the inverse scattering problem, where an initial guess for the optimization problem is obtained using a trained neural network.
no code implementations • 7 Oct 2022 • Xingyu Zhu, Zixuan Wang, Xiang Wang, Mo Zhou, Rong Ge
Globally we observe that the training dynamics for our example has an interesting bifurcating behavior, which was also observed in the training of neural nets.
no code implementations • 3 Oct 2022 • Xiang Wang, Annie N. Wang, Mo Zhou, Rong Ge
Monotonic linear interpolation (MLI) - on the line connecting a random initialization with the minimizer it converges to, the loss and accuracy are monotonic - is a phenomenon that is commonly observed in the training of neural networks.
no code implementations • 19 May 2022 • Mo Zhou, Vishal M. Patel
Adversarial attacks pose safety and security concerns to deep learning applications, but their characteristics are under-explored.
2 code implementations • CVPR 2022 • Mo Zhou, Vishal M. Patel
Owing to security implications of adversarial vulnerability, adversarial robustness of deep metric learning models has to be improved.
no code implementations • 2 Feb 2022 • Zeping Luo, Shiyou Wu, Cindy Weng, Mo Zhou, Rong Ge
Self-supervised learning has significantly improved the performance of many NLP tasks.
no code implementations • 31 Jan 2022 • Mo Zhou, Jianfeng Lu
We propose a single time-scale actor-critic algorithm to solve the linear quadratic regulator (LQR) problem.
no code implementations • 29 Sep 2021 • Zeping Luo, Cindy Weng, Shiyou Wu, Mo Zhou, Rong Ge
Self-supervised learning has significantly improved the performance of many NLP tasks.
no code implementations • CVPR 2021 • Liushuai Shi, Le Wang, Chengjiang Long, Sanping Zhou, Mo Zhou, Zhenxing Niu, Gang Hua
Specifically, the SGCN explicitly models the sparse directed interaction with a sparse directed spatial graph to capture adaptive interaction pedestrians.
no code implementations • NeurIPS 2021 • Rong Ge, Yunwei Ren, Xiang Wang, Mo Zhou
In this paper we study the training dynamics for gradient flow on over-parametrized tensor decomposition problems.
1 code implementation • 7 Jun 2021 • Mo Zhou, Le Wang, Zhenxing Niu, Qilin Zhang, Nanning Zheng, Gang Hua
In this paper, we propose two attacks against deep ranking systems, i. e., Candidate Attack and Query Attack, that can raise or lower the rank of chosen candidates by adversarial perturbations.
4 code implementations • 4 Apr 2021 • Liushuai Shi, Le Wang, Chengjiang Long, Sanping Zhou, Mo Zhou, Zhenxing Niu, Gang Hua
Meanwhile, we use a sparse directed temporal graph to model the motion tendency, thus to facilitate the prediction based on the observed direction.
2 code implementations • ICCV 2021 • Mo Zhou, Le Wang, Zhenxing Niu, Qilin Zhang, Yinghui Xu, Nanning Zheng, Gang Hua
In this paper, we formulate a new adversarial attack against deep ranking systems, i. e., the Order Attack, which covertly alters the relative order among a selected set of candidates according to an attacker-specified permutation, with limited interference to other unrelated candidates.
no code implementations • 4 Feb 2021 • Mo Zhou, Rong Ge, Chi Jin
We show that as long as the loss is already lower than a threshold (polynomial in relevant parameters), all student neurons in an over-parameterized two-layer neural network will converge to one of teacher neurons, and the loss will go to 0.
no code implementations • 1 Jan 2021 • Mo Zhou, Le Wang, Zhenxing Niu, Qilin Zhang, Xu Yinghui, Nanning Zheng, Gang Hua
The objective of this paper is to formalize and practically implement a new adversarial attack against deep ranking systems, i. e., the Order Attack, which covertly alters the relative order of a selected set of candidates according to a permutation vector predefined by the attacker, with only limited interference to other unrelated candidates.
3 code implementations • ECCV 2020 • Mo Zhou, Zhenxing Niu, Le Wang, Qilin Zhang, Gang Hua
In this paper, we propose two attacks against deep ranking systems, i. e., Candidate Attack and Query Attack, that can raise or lower the rank of chosen candidates by adversarial perturbations.
no code implementations • 7 Feb 2020 • Jiequn Han, Jianfeng Lu, Mo Zhou
We propose a new method to solve eigenvalue problems for linear and semilinear second order differential operators in high dimensions based on deep neural networks.
2 code implementations • 18 Nov 2019 • Mo Zhou, Zhenxing Niu, Le Wang, Zhanning Gao, Qilin Zhang, Gang Hua
For visual-semantic embedding, the existing methods normally treat the relevance between queries and candidates in a bipolar way -- relevant or irrelevant, and all "irrelevant" candidates are uniformly pushed away from the query by an equal margin in the embedding space, regardless of their various proximity to the query.
no code implementations • NeurIPS 2019 • Tianyi Liu, Minshuo Chen, Mo Zhou, Simon S. Du, Enlu Zhou, Tuo Zhao
We show, however, that gradient descent combined with proper normalization, avoids being trapped by the spurious local optimum, and converges to a global optimum in polynomial time, when the weight of the first layer is initialized at 0, and that of the second layer is initialized arbitrarily in a ball.
no code implementations • 7 Sep 2019 • Mo Zhou, Tianyi Liu, Yan Li, Dachao Lin, Enlu Zhou, Tuo Zhao
Numerous empirical evidence has corroborated that the noise plays a crucial rule in effective and efficient training of neural networks.
no code implementations • ICCV 2017 • Zhenxing Niu, Mo Zhou, Le Wang, Xinbo Gao, Gang Hua
We address the problem of dense visual-semantic embedding that maps not only full sentences and whole images but also phrases within sentences and salient regions within images into a multimodal embedding space.
no code implementations • CVPR 2016 • Zhenxing Niu, Mo Zhou, Le Wang, Xinbo Gao, Gang Hua
To address the non-stationary property of aging patterns, age estimation can be cast as an ordinal regression problem.