Efficient, Scalable, and Sustainable DNN Training on SoC-Clustered Edge Servers.
IEEE Trans. Mob. Comput., December, 2024
ELMS: Elasticized Large Language Models On Mobile Devices.
CoRR, 2024
Empowering 1000 tokens/second on-device LLM prefilling with mllm-NPU.
CoRR, 2024
A Survey of Resource-efficient LLM and Multimodal Foundation Models.
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
CoRR, 2024
Towards Energy-efficient Federated Learning via INT8-based Training on Mobile DSPs.
Proceedings of the ACM on Web Conference 2024, 2024
PieBridge: Fast and Parameter-Efficient On-Device Training via Proxy Networks.
Proceedings of the 22nd ACM Conference on Embedded Networked Sensor Systems, 2024
WiP: Efficient LLM Prefilling with Mobile NPU.
Proceedings of the Workshop on Edge and Mobile Foundation Models, 2024
SoCFlow: Efficient and Scalable DNN Training on SoC-Clustered Edge Servers.
Proceedings of the 29th ACM International Conference on Architectural Support for Programming Languages and Operating Systems, 2024
LLMCad: Fast and Scalable On-device Large Language Model Inference.
CoRR, 2023
Niagara: Scheduling DNN Inference Services on Heterogeneous Edge Processors.
Proceedings of the Service-Oriented Computing - 21st International Conference, 2023
Satellite Computing: From Space to Your Screen.
Proceedings of the Service-Oriented Computing - ICSOC 2023 Workshops - AI-PA, ASOCA, SAPD, SQS, SSCOPE, WESOACS and Satellite Events, Rome, Italy, November 28, 2023
Mandheling: Mixed-Precision On-Device DNN Training with DSP Offloading.
CoRR, 2022
Mandheling: mixed-precision on-device DNN training with DSP offloading.
Proceedings of the ACM MobiCom '22: The 28th Annual International Conference on Mobile Computing and Networking, Sydney, NSW, Australia, October 17, 2022
S3Library: Automatically Eliminating C/C++ Buffer Overflow using Compatible Safer Libraries.
CoRR, 2020
DangKiller: Eliminating Dangling Pointers Efficiently via Implicit Identifier.
CoRR, 2020
SMA: Eliminate Memory Spatial Errors via Saturation Memory Access.
CoRR, 2020
An adaptive template matching-based single object tracking algorithm with parallel acceleration.
J. Vis. Commun. Image Represent., 2019