Search Results for author:
Found 5 papers, 1 papers with code
Enabling On-Device Large Language Model Personalization with Self-Supervised Data Selection and Synthesis
While it is possible to obtain annotation locally by directly asking users to provide preferred responses, such annotations have to be sparse to not affect user experience.
Enabling Weakly-Supervised Temporal Action Localization from On-Device Learning of the Video Stream
To enable W-TAL models to learn from a long, untrimmed streaming video, we propose an efficient video learning approach that can directly adapt to new environments.
Action Detection
Weakly-supervised Temporal Action Localization
+1
Sustainable AI Processing at the Edge
In particular, we explore the use of processing-in-memory (PIM) approaches, mobile GPU accelerators, and recently released FPGAs, and compare them with novel Racetrack memory PIM.
H2H: Heterogeneous Model to Heterogeneous System Mapping with Computation and Communication Awareness
The heterogeneity in ML models comes from multi-sensor perceiving and multi-task learning, i. e., multi-modality multi-task (MMMT), resulting in diverse deep neural network (DNN) layers and computation patterns.
EF-Train: Enable Efficient On-device CNN Training on FPGA Through Data Reshaping for Online Adaptation or Personalization
In this work, we design EF-Train, an efficient DNN training accelerator with a unified channel-level parallelism-based convolution kernel that can achieve end-to-end training on resource-limited low-power edge-level FPGAs.
