Search Results for author: Kaidi Cao

Found 25 papers, 13 papers with code

GraphMETRO: Mitigating Complex Graph Distribution Shifts via Mixture of Aligned Experts

no code implementations7 Dec 2023 Shirley Wu, Kaidi Cao, Bruno Ribeiro, James Zou, Jure Leskovec

Graph data are inherently complex and heterogeneous, leading to a high natural diversity of distributional shifts.

TpuGraphs: A Performance Prediction Dataset on Large Tensor Computational Graphs

1 code implementation NeurIPS 2023 Phitchaya Mangpo Phothilimthana, Sami Abu-El-Haija, Kaidi Cao, Bahare Fatemi, Mike Burrows, Charith Mendis, Bryan Perozzi

TpuGraphs provides 25x more graphs than the largest graph property prediction dataset (with comparable graph sizes), and 770x larger graphs on average compared to existing performance prediction datasets on machine learning programs.

Graph Property Prediction Property Prediction

Communication-Free Distributed GNN Training with Vertex Cut

no code implementations6 Aug 2023 Kaidi Cao, Rui Deng, Shirley Wu, Edward W Huang, Karthik Subbian, Jure Leskovec

Here, we introduce CoFree-GNN, a novel distributed GNN training framework that significantly speeds up the training process by implementing communication-free training.

Learning Large Graph Property Prediction via Graph Segment Training

1 code implementation NeurIPS 2023 Kaidi Cao, Phitchaya Mangpo Phothilimthana, Sami Abu-El-Haija, Dustin Zelle, Yanqi Zhou, Charith Mendis, Jure Leskovec, Bryan Perozzi

Here we propose Graph Segment Training (GST), a general framework that utilizes a divide-and-conquer approach to allow learning large graph property prediction with a constant memory footprint.

Graph Property Prediction Property Prediction

AutoTransfer: AutoML with Knowledge Transfer -- An Application to Graph Neural Networks

1 code implementation14 Mar 2023 Kaidi Cao, Jiaxuan You, Jiaju Liu, Jure Leskovec

Experiments demonstrate that (i) our proposed task embedding can be computed efficiently, and that tasks with similar embeddings have similar best-performing architectures; (ii) AutoTransfer significantly improves search efficiency with the transferred design priors, reducing the number of explored architectures by an order of magnitude.

AutoML Transfer Learning

Relational Multi-Task Learning: Modeling Relations between Data and Tasks

1 code implementation ICLR 2022 Kaidi Cao, Jiaxuan You, Jure Leskovec

Here we introduce a novel relational multi-task learning setting where we leverage data point labels from auxiliary tasks to make more accurate predictions on the new task.

Multi-Task Learning

TuneUp: A Simple Improved Training Strategy for Graph Neural Networks

no code implementations26 Oct 2022 Weihua Hu, Kaidi Cao, Kexin Huang, Edward W Huang, Karthik Subbian, Kenji Kawaguchi, Jure Leskovec

Extensive evaluation of TuneUp on five diverse GNN architectures, three types of prediction tasks, and both transductive and inductive settings shows that TuneUp significantly improves the performance of the base GNN on tail nodes, while often even improving the performance on head nodes.

Data Augmentation

Learning Large-scale Subsurface Simulations with a Hybrid Graph Network Simulator

no code implementations15 Jun 2022 Tailin Wu, Qinchen Wang, Yinan Zhang, Rex Ying, Kaidi Cao, Rok Sosič, Ridwan Jalali, Hassan Hamam, Marko Maucec, Jure Leskovec

To model complex reservoir dynamics at both local and global scale, HGNS consists of a subsurface graph neural network (SGNN) to model the evolution of fluid flows, and a 3D-U-Net to model the evolution of pressure.

Decision Making

Learning Backward Compatible Embeddings

1 code implementation7 Jun 2022 Weihua Hu, Rajas Bansal, Kaidi Cao, Nikhil Rao, Karthik Subbian, Jure Leskovec

We formalize the problem where the goal is for the embedding team to keep updating the embedding version, while the consumer teams do not have to retrain their models.

Fraud Detection Product Recommendation +1

Open-World Semi-Supervised Learning

1 code implementation ICLR 2022 Kaidi Cao, Maria Brbic, Jure Leskovec

Here, we introduce a novel open-world semi-supervised learning setting that formalizes the notion that novel classes may appear in the unlabeled test data.

Image Classification Novel Object Detection +1

Coresets for Robust Training of Deep Neural Networks against Noisy Labels

no code implementations NeurIPS 2020 Baharan Mirzasoleiman, Kaidi Cao, Jure Leskovec

Modern neural networks have the capacity to overfit noisy labels frequently found in real-world datasets.

Concept Learners for Few-Shot Learning

2 code implementations ICLR 2021 Kaidi Cao, Maria Brbic, Jure Leskovec

Developing algorithms that are able to generalize to a novel task given only a few labeled examples represents a fundamental challenge in closing the gap between machine- and human-level performance.

Few-Shot Learning Fine-Grained Image Classification

Heteroskedastic and Imbalanced Deep Learning with Adaptive Regularization

1 code implementation ICLR 2021 Kaidi Cao, Yining Chen, Junwei Lu, Nikos Arechiga, Adrien Gaidon, Tengyu Ma

Real-world large-scale datasets are heteroskedastic and imbalanced -- labels have varying levels of uncertainty and label distributions are long-tailed.

Image Classification

Learning Temporal Action Proposals With Fewer Labels

no code implementations ICCV 2019 Jingwei Ji, Kaidi Cao, Juan Carlos Niebles

Most current methods for training action proposal modules rely on fully supervised approaches that require large amounts of annotated temporal action intervals in long video sequences.

Action Detection Semi-Supervised Action Detection

Delving Deep Into Hybrid Annotations for 3D Human Recovery in the Wild

1 code implementation ICCV 2019 Yu Rong, Ziwei Liu, Cheng Li, Kaidi Cao, Chen Change Loy

Specifically, we focus on the challenging task of in-the-wild 3D human recovery from single images when paired 3D annotations are not fully available.

Learning Imbalanced Datasets with Label-Distribution-Aware Margin Loss

7 code implementations NeurIPS 2019 Kaidi Cao, Colin Wei, Adrien Gaidon, Nikos Arechiga, Tengyu Ma

Deep learning algorithms can fare poorly when the training dataset suffers from heavy class-imbalance but the testing criterion requires good generalization on less frequent classes.

Long-tail learning with class descriptors

Disentangling Content and Style via Unsupervised Geometry Distillation

1 code implementation ICLR Workshop DeepGenStruct 2019 Wayne Wu, Kaidi Cao, Cheng Li, Chen Qian, Chen Change Loy

It is challenging to disentangle an object into two orthogonal spaces of content and style since each can influence the visual observation differently and unpredictably.

Disentanglement

TransGaGa: Geometry-Aware Unsupervised Image-to-Image Translation

no code implementations CVPR 2019 Wayne Wu, Kaidi Cao, Cheng Li, Chen Qian, Chen Change Loy

Extensive experiments demonstrate the superior performance of our method to other state-of-the-art approaches, especially in the challenging near-rigid and non-rigid objects translation tasks.

Translation Unsupervised Image-To-Image Translation

CariGANs: Unpaired Photo-to-Caricature Translation

no code implementations1 Nov 2018 Kaidi Cao, Jing Liao, Lu Yuan

Facial caricature is an art form of drawing faces in an exaggerated way to convey humor or sarcasm.

Caricature Generative Adversarial Network +2

Unsupervised Disentangling Structure and Appearance

no code implementations27 Sep 2018 Wayne Wu, Kaidi Cao, Cheng Li, Chen Qian, Chen Change Loy

It is challenging to disentangle an object into two orthogonal spaces of structure and appearance since each can influence the visual observation in a different and unpredictable way.

Disentanglement

DNN Dataflow Choice Is Overrated

no code implementations10 Sep 2018 Xuan Yang, Mingyu Gao, Jing Pu, Ankita Nayak, Qiaoyi Liu, Steven Emberton Bell, Jeff Ou Setter, Kaidi Cao, Heonjae Ha, Christos Kozyrakis, Mark Horowitz

Many DNN accelerators have been proposed and built using different microarchitectures and program mappings.

Distributed, Parallel, and Cluster Computing

Pose-Robust Face Recognition via Deep Residual Equivariant Mapping

1 code implementation CVPR 2018 Kaidi Cao, Yu Rong, Cheng Li, Xiaoou Tang, Chen Change Loy

However, many contemporary face recognition models still perform relatively poor in processing profile faces compared to frontal faces.

Face Identification Face Recognition +2

Merge or Not? Learning to Group Faces via Imitation Learning

1 code implementation13 Jul 2017 Yue He, Kaidi Cao, Cheng Li, Chen Change Loy

Given a large number of unlabeled face images, face grouping aims at clustering the images into individual identities present in the data.

Clustering Imitation Learning

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