ETRI-Knowledge Sharing Plaform

ENGLISH

성과물

논문 검색
구분 SCI
연도 ~ 키워드

상세정보

학술대회 mobile YOLACT: Toward Lightweight Instance Segmentation for Mobile Devices
Cited 3 time in scopus Download 2 time Share share facebook twitter linkedin kakaostory
저자
이주원, 이승재, 고종국
발행일
202110
출처
International Conference on Information and Communication Technology Convergence (ICTC) 2021, pp.1456-1460
DOI
https://dx.doi.org/10.1109/ICTC52510.2021.9621125
협약과제
21HH5100, 객체추출 및 실-가상 정합 지원 모바일 AR 기술 개발, 고종국
초록
In this paper, we present a lightweight instance segmentation model, mobileYOLACT which is designed for mobile environments where the computational resources are limited. We propose several modifications to YOLACT to improve computational efficiency. First, we use a quantized lightweight backbone for feature extraction. Second, we reduce the computational burden with marginal degradation in accuracy by employing the depthwise separable convolution on the entire model. Third, we simplified the structure of prototype mask generation branch. Last, we used TorchScript and NCNN to further optimize the model and deploy it on mobile device. We validate the effectiveness of the proposed method from the experiments on COCO dataset. The proposed model can run at the speed of 21 FPS on Samsung Galaxy S20 with 23 APmask at 0.5 IoU threshold.
KSP 제안 키워드
Computational Efficiency, Feature extractioN, Mask generation, Mobile devices, Proposed model, Samsung Galaxy S2, computational burden, computational resources, mobile environment