Abstract | ||
---|---|---|
Over the past years, feed-forward convolutional neural networks (CNNs) have evolved from a simple feed-forward architecture to deep and residual (skip-connection) architectures, demonstrating increasingly higher object categorization accuracy and increasingly better explanatory power of both neural and behavioral responses. However, from the neuroscientist point of view, the relationship between such deep architectures and the ventral visual pathway is incomplete. For example, current state-of-the-art CNNs appear to be too complex (e.g., now over 100 layers for ResNet) compared with the relatively shallow cortical hierarchy (4-8 layers). We introduce new CNNs with shallow recurrent architectures and skip connections requiring fewer parameters. With higher accuracy for classification, we propose an architecture for recurrent residual convolutional neural network (R2CNN) on FPGA, which efficiently utilizes on-chip memory bandwidth. We propose an Output-Kernel- Input-Parallel (OKIP) convolution circuit for a recurrent residual convolution stage. We implement the inference hardware on a Xilinx ZCU104 evaluation board with high-level synthesis. Our R2CNN accelerator achieves top-5 accuracy of 90.08% on ImageNet bench- mark, which has higher accuracy than conventional FPGA implementations.
|
Year | DOI | Venue |
---|---|---|
2020 | 10.1145/3373087.3375367 | FPGA |
Field | DocType | ISBN |
Residual,Computer science,Convolutional neural network,Parallel computing,Field-programmable gate array | Conference | 978-1-4503-7099-8 |
Citations | PageRank | References |
0 | 0.34 | 0 |
Authors | ||
4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Hiroki Nakahara | 1 | 155 | 37.34 |
Zhiqiang Que | 2 | 26 | 9.81 |
Akira Jinguji | 3 | 5 | 4.18 |
Wayne Luk | 4 | 3752 | 438.09 |