FMTT: Fused Multi-Head Transformer with Tensor-Compression for 3D Point Clouds Detection on Edge Devices

Zikun Wei; Tingting Wang; Chenchen Ding; Bohan Wang; Ziyi Guan; Hantao Huang; Hao Yu

2024-03-27

1530-1591

979-8-3503-4860-6

2024 Design, Automation & Test in Europe Conference & Exhibition (DATE)

25-27 March 2024

Valencia, Spain

The real-time detection of 3D objects represents a grand challenge on edge devices. Existing 3D point clouds models are over-parameterized with heavy computation load. This paper proposes a highly compact model for 3D point clouds detection using tensor-compression. Compared to conventional methods, we propose a fused multi-head transformer tensor-compression (FMTT) to achieve both compact size yet with high accuracy. The FMTT leverages different ranks to extract both high and low-level features and then fuses them together to improve the accuracy. Experiments on the KITTI dataset show that the proposed FMTT can achieve 6.04× smaller than the uncompressed model from 55.09MB to 9.12MB such that the compressed model can be implemented on edge devices. It also achieves 2.62% improved accuracy in easy mode and 0.28% improved accuracy in hard mode.

第一

EI

20242516298234

3D modeling ; Deep learning ; Object detection ; Object recognition

Ergonomics and Human Factors Engineering:461.4 ; Data Processing and Image Processing:723.2 ; Algebra:921.1

School of Microelectronics, Southern University of Science and Technology, Shenzhen, China

Zikun Wei,Tingting Wang,Chenchen Ding,et al. FMTT: Fused Multi-Head Transformer with Tensor-Compression for 3D Point Clouds Detection on Edge Devices[C],2024.