papers
Publications (2)
cs.CV2015
Deep Image: Scaling up Image Recognition
Ren Wu, Shengen Yan, Yi Shan +2
cs.DC2019
DNNVM : End-to-End Compiler Leveraging Heterogeneous Optimizations on FPGA-based CNN Accelerators
Yu Xing, Shuang Liang, Lingzhi Sui +6
DNNVM is a full‑stack compiler that optimizes graph, loop, and data‑layout aspects of CNN models for FPGA‑based accelerators, improving throughput and energy efficiency through het…
#fpga acceleration#cnn compilation#graph optimization#data layout