Easy2Hard: From Partially to Fully Unmatched Modalities as Negative Samples in Contrastive Learning
Contrastive learning
Multimodel
CVPR 2026
Multimodal Contrasive learning
AgenticWorld
Agentic
AgenticWorld
World
Lab
Lab
Lab
We develop the foundations of agentic AI: world models, embodied reasoning, and autonomous systems that understand and operate in the physical world.
Federated Agent
World action model
Physical AI
Visual agent
Easy2Hard: From Partially to Fully Unmatched Modalities as Negative Samples in Contrastive Learning
Contrastive learning
Multimodel
CVPR 2026
Multimodal Contrasive learning
Easy2Hard: From Partially to Fully Unmatched Modalities as Negative Samples in Contrastive Learning
Contrastive learning
Multimodel
CVPR 2026
Multimodal Contrasive learning
FLSys: Toward an Open Ecosystem for Federated Learning Mobile Apps
Federate learning
Edge devices
IEEE Transactions on Mobile Computing
Distributed federate learning
FLSys: Toward an Open Ecosystem for Federated Learning Mobile Apps
Federate learning
Edge devices
IEEE Transactions on Mobile Computing
Distributed federate learning
FLSys: Toward an Open Ecosystem for Federated Learning Mobile Apps
Federate learning
Edge devices
IEEE Transactions on Mobile Computing
Distributed federate learning
Complement Sparsification: Low-Overhead Model Pruning for Federated Learning
Model Pruning
Federate learning
AAAI 2023
Federated learning
Complement Sparsification: Low-Overhead Model Pruning for Federated Learning
Model Pruning
Federate learning
AAAI 2023
Federated learning
Complement Sparsification: Low-Overhead Model Pruning for Federated Learning
Model Pruning
Federate learning
AAAI 2023
Federated learning