Search Results for author: Dongxiao Zhu

Found 32 papers, 13 papers with code

Learning to Poison Large Language Models During Instruction Tuning

1 code implementation21 Feb 2024 Yao Qiang, Xiangyu Zhou, Saleh Zare Zade, Mohammad Amin Roshani, Douglas Zytko, Dongxiao Zhu

The advent of Large Language Models (LLMs) has marked significant achievements in language processing and reasoning capabilities.

Data Poisoning

MFABA: A More Faithful and Accelerated Boundary-based Attribution Method for Deep Neural Networks

1 code implementation21 Dec 2023 Zhiyu Zhu, Huaming Chen, Jiayu Zhang, Xinyi Wang, Zhibo Jin, Minhui Xue, Dongxiao Zhu, Kim-Kwang Raymond Choo

To better understand the output of deep neural networks (DNN), attribution based methods have been an important approach for model interpretability, which assign a score for each input dimension to indicate its importance towards the model outcome.

FedDRO: Federated Compositional Optimization for Distributionally Robust Learning

no code implementations21 Nov 2023 Prashant Khanduri, Chengyin Li, Rafi Ibn Sultan, Yao Qiang, Joerg Kliewer, Dongxiao Zhu

A key novelty of our work is to develop solution accuracy-independent algorithms that do not require large batch gradients (and function evaluations) for solving federated CO problems.

Federated Learning

Hijacking Large Language Models via Adversarial In-Context Learning

1 code implementation16 Nov 2023 Yao Qiang, Xiangyu Zhou, Dongxiao Zhu

In-context learning (ICL) has emerged as a powerful paradigm leveraging LLMs for specific tasks by utilizing labeled examples as demonstrations in the precondition prompts.

In-Context Learning Specificity

Interpretability-Aware Vision Transformer

1 code implementation14 Sep 2023 Yao Qiang, Chengyin Li, Prashant Khanduri, Dongxiao Zhu

Furthermore, if ViTs are not properly trained with the given data and do not prioritize the region of interest, the {\it post hoc} methods would be less effective.

Image Classification

Auto-Prompting SAM for Mobile Friendly 3D Medical Image Segmentation

no code implementations28 Aug 2023 Chengyin Li, Prashant Khanduri, Yao Qiang, Rafi Ibn Sultan, Indrin Chetty, Dongxiao Zhu

In addition to the domain gaps between natural and medical images, disparities in the spatial arrangement between 2D and 3D images, the substantial computational burden imposed by powerful GPU servers, and the time-consuming manual prompt generation impede the extension of SAM to a broader spectrum of medical image segmentation applications.

Image Segmentation Medical Image Segmentation +3

Fairness-aware Vision Transformer via Debiased Self-Attention

no code implementations31 Jan 2023 Yao Qiang, Chengyin Li, Prashant Khanduri, Dongxiao Zhu

Importantly, our DSA framework leads to improved fairness guarantees over prior works on multiple prediction tasks without compromising target prediction performance.

Fairness

Negative Flux Aggregation to Estimate Feature Attributions

1 code implementation17 Jan 2023 Xin Li, Deng Pan, Chengyin Li, Yao Qiang, Dongxiao Zhu

There are increasing demands for understanding deep neural networks' (DNNs) behavior spurred by growing security and/or transparency concerns.

Learning Compact Features via In-Training Representation Alignment

no code implementations23 Nov 2022 Xin Li, Xiangrui Li, Deng Pan, Yao Qiang, Dongxiao Zhu

Deep neural networks (DNNs) for supervised learning can be viewed as a pipeline of the feature extractor (i. e., last hidden layer) and a linear classifier (i. e., output layer) that are trained jointly with stochastic gradient descent (SGD) on the loss function (e. g., cross-entropy).

Representation Learning

Coupling User Preference with External Rewards to Enable Driver-centered and Resource-aware EV Charging Recommendation

1 code implementation23 Oct 2022 Chengyin Li, Zheng Dong, Nathan Fisher, Dongxiao Zhu

Electric Vehicle (EV) charging recommendation that both accommodates user preference and adapts to the ever-changing external environment arises as a cost-effective strategy to alleviate the range anxiety of private EV drivers.

FocalUNETR: A Focal Transformer for Boundary-aware Segmentation of CT Images

1 code implementation6 Oct 2022 Chengyin Li, Yao Qiang, Rafi Ibn Sultan, Hassan Bagher-Ebadian, Prashant Khanduri, Indrin J. Chetty, Dongxiao Zhu

Computed Tomography (CT) based precise prostate segmentation for treatment planning is challenging due to (1) the unclear boundary of the prostate derived from CT's poor soft tissue contrast and (2) the limitation of convolutional neural network-based models in capturing long-range global context.

Computed Tomography (CT) Image Segmentation +2

Saliency Guided Adversarial Training for Learning Generalizable Features with Applications to Medical Imaging Classification System

no code implementations9 Sep 2022 Xin Li, Yao Qiang, Chengyin Li, Sijia Liu, Dongxiao Zhu

We hypothesize that adversarial training can eliminate shortcut features whereas saliency guided training can filter out non-relevant features; both are nuisance features accounting for the performance degradation on OOD test sets.

Adversarially Robust and Explainable Model Compression with On-Device Personalization for Text Classification

no code implementations10 Jan 2021 Yao Qiang, Supriya Tumkur Suresh Kumar, Marco Brocanelli, Dongxiao Zhu

On-device Deep Neural Networks (DNNs) have recently gained more attention due to the increasing computing power of the mobile devices and the number of applications in Computer Vision (CV), Natural Language Processing (NLP), and Internet of Things (IoTs).

Adversarial Robustness General Classification +4

Improving Adversarial Robustness via Probabilistically Compact Loss with Logit Constraints

1 code implementation14 Dec 2020 Xin Li, Xiangrui Li, Deng Pan, Dongxiao Zhu

This inspires us to propose a new Probabilistically Compact (PC) loss with logit constraints which can be used as a drop-in replacement for cross-entropy (CE) loss to improve CNN's adversarial robustness.

Adversarial Robustness

Explainable Recommendation via Interpretable Feature Mapping and Evaluation of Explainability

no code implementations12 Jul 2020 Deng Pan, Xiangrui Li, Xin Li, Dongxiao Zhu

Latent factor collaborative filtering (CF) has been a widely used technique for recommender system by learning the semantic representations of users and items.

Collaborative Filtering Explainable Recommendation +1

Defending against adversarial attacks on medical imaging AI system, classification or detection?

1 code implementation24 Jun 2020 Xin Li, Deng Pan, Dongxiao Zhu

Medical imaging AI systems such as disease classification and segmentation are increasingly inspired and transformed from computer vision based AI systems.

Adversarial Defense General Classification

COVID-MobileXpert: On-Device COVID-19 Patient Triage and Follow-up using Chest X-rays

1 code implementation6 Apr 2020 Xin Li, Chengyin Li, Dongxiao Zhu

We design and implement a novel three-player knowledge transfer and distillation (KTD) framework including a pre-trained attending physician (AP) network that extracts CXR imaging features from a large scale of lung disease CXR images, a fine-tuned resident fellow (RF) network that learns the essential CXR imaging features to discriminate COVID-19 from pneumonia and/or normal cases with a small amount of COVID-19 cases, and a trained lightweight medical student (MS) network to perform on-device COVID-19 patient triage and follow-up.

Computed Tomography (CT) Trajectory Prediction +1

Toward Tag-free Aspect Based Sentiment Analysis: A Multiple Attention Network Approach

3 code implementations22 Mar 2020 Yao Qiang, Xin Li, Dongxiao Zhu

Existing aspect based sentiment analysis (ABSA) approaches leverage various neural network models to extract the aspect sentiments via learning aspect-specific feature representations.

Aspect-Based Sentiment Analysis Aspect-Based Sentiment Analysis (ABSA) +1

On the Learning Property of Logistic and Softmax Losses for Deep Neural Networks

1 code implementation4 Mar 2020 Xiangrui Li, Xin Li, Deng Pan, Dongxiao Zhu

Deep convolutional neural networks (CNNs) trained with logistic and softmax losses have made significant advancement in visual recognition tasks in computer vision.

Binary Classification Classification +2

Improve SGD Training via Aligning Mini-batches

no code implementations23 Feb 2020 Xiangrui Li, Deng Pan, Xin Li, Dongxiao Zhu

In each iteration of SGD, a mini-batch from the training data is sampled and the true gradient of the loss function is estimated as the noisy gradient calculated on this mini-batch.

Representation Learning with Autoencoders for Electronic Health Records: A Comparative Study

no code implementations24 Aug 2019 Najibesadat Sadati, Milad Zafar Nezhad, Ratna Babu Chinnam, Dongxiao Zhu

Our focus is to present a comparative study to evaluate the performance of different deep architectures through supervised learning and provide insights in the choice of deep feature representation techniques.

Representation Learning Small Data Image Classification

Vispi: Automatic Visual Perception and Interpretation of Chest X-rays

no code implementations MIDL 2019 Xin Li, Rui Cao, Dongxiao Zhu

Medical imaging contains the essential information for rendering diagnostic and treatment decisions.

Image Captioning

Interpreting Age Effects of Human Fetal Brain from Spontaneous fMRI using Deep 3D Convolutional Neural Networks

no code implementations9 Jun 2019 Xiangrui Li, Jasmine Hect, Moriah Thomason, Dongxiao Zhu

The findings demonstrate that deep CNNs are a promising approach for identifying spontaneous functional patterns in fetal brain activity that discriminate age groups.

Multi-task Prediction of Patient Workload

no code implementations27 Dec 2018 Mohammad Hessam Olya, Dongxiao Zhu, Kai Yang

This issue becomes more critical for the healthcare facilities that provide service for chronic disease treatment because of the need for continuous treatments over the time.

Decision Making Multi-Task Learning

Representation Learning with Autoencoders for Electronic Health Records: A Comparative Study

no code implementations6 Jan 2018 Najibesadat Sadati, Milad Zafar Nezhad, Ratna Babu Chinnam, Dongxiao Zhu

Our focus is to present a comparative study to evaluate the performance of different deep architectures through supervised learning and provide insights in the choice of deep feature representation techniques.

Representation Learning Small Data Image Classification

SUBIC: A Supervised Bi-Clustering Approach for Precision Medicine

no code implementations26 Sep 2017 Milad Zafar Nezhad, Dongxiao Zhu, Najibesadat Sadati, Kai Yang, Phillip Levy

Traditional medicine typically applies one-size-fits-all treatment for the entire patient population whereas precision medicine develops tailored treatment schemes for different patient subgroups.

Clustering

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