Name
Abstract | ||
|---|---|---|
Depth sensing is crucial for understanding 3D scenes on embedded systems such as home robots, self-driving cars, and drones. Monocular depth estimation which gives pixel-wise depth from a general camera, has attracted attention in recent years, due to the reliability, low-cost and small area requirement. Past research by using Convolutional Neural Network (CNN) has gained high accuracy and been increasing interest. However, the CNN requires a massive amount of MACs (Multiply ACcumulations) and weights, so its latency is extremely long. To address this problem, we present hardware-oriented pruning for separable convolutions and effectively parallelized MAC Unit. We introduce a filter-wise pruned DepthFCN and novel FPGA architecture that exploit its sparsity. Moreover, dense convolution and pruned separable convolution are implemented on a shared convolutional circuit due to high hardware efficiency and a high parallel degree. We compare the proposed FPGA-based system with the Jetson TX2. The FPGA accelerator achieves 123.6 FPS with 0.3 W power consumption for a 256×256 image, and its accuracy is 76.2%. Compared with the mobile GPU, it is 1.5 times faster and its power consumption is 20 times lower. We demonstrate the fastest monocular depth estimation by using a low-cost FPGA board that is suitable for embedded systems. |
Year | DOI | Venue |
|---|---|---|
2020 | 10.1109/IPDPSW50202.2020.00032 | 2020 IEEE International Parallel and Distributed Processing Symposium Workshops (IPDPSW) |
Keywords | DocType | ISSN |
Jetson TX2,low-cost FPGA board,embedded systems,depth sensing,understanding 3D scenes,home robots,self-driving cars,drones,convolutional neural network,CNN,MAC unit,filter-wise pruned DepthFCN,pruned separable convolution,shared convolutional circuit,FPGA accelerator system,monocular depth estimation,multiply accumulations,power 0.3 W | Conference | 2164-7062 |
ISBN | Citations | PageRank |
978-1-7281-7457-0 | 0 | 0.34 |
References | Authors | |
2 | 6 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Youki Sada | 1 | 1 | 1.79 |
| Naoto Soga | 2 | 0 | 0.34 |
| Masayuki Shimoda | 3 | 8 | 6.45 |
| Akira Jinguji | 4 | 5 | 4.18 |
| Shimpei Sato | 5 | 43 | 13.03 |
| Hiroki Nakahara | 6 | 155 | 37.34 |