인문학
사회과학
자연과학
공학
의약학
농수해양학
예술체육학
복합학
지원사업
학술연구/단체지원/교육 등 연구자 활동을 지속하도록 DBpia가 지원하고 있어요.
커뮤니티
연구자들이 자신의 연구와 전문성을 널리 알리고, 새로운 협력의 기회를 만들 수 있는 네트워킹 공간이에요.
초록·키워드
Besides the well-known challenges of object tracking, and the amazing improvements in visual tracking robustness, many tracking algorithms failed to take into strong consideration of the accuracy of tracking. In our paper we highlighted ways not only to overcome issues of target objects which undergo significant motion and appearance changes, but also to enhance the quality of features, which are the core ingredients of major deep neural network algorithms.
In this work, we use features of VGG19 pretrained on ImageNet dataset and propose a novel Spatial-Semantic Residual Features for Object Tracking (SSReF) model to enhance the quality of features and the flow of the gradient by taking advantage of both earlier and later layer features of the network, and by incorporating residual features. The experimental results show favorable tracking accuracy and speed.
In this work, we use features of VGG19 pretrained on ImageNet dataset and propose a novel Spatial-Semantic Residual Features for Object Tracking (SSReF) model to enhance the quality of features and the flow of the gradient by taking advantage of both earlier and later layer features of the network, and by incorporating residual features. The experimental results show favorable tracking accuracy and speed.
본문·목차
인공지능 문자 인식 모델을 통해 추출된 텍스트로, 일부 오타나 오류가 포함될 수 있으나 지속적으로 개선 중입니다.
오류를 발견하셨다면 해당 부분을 드래그한 후 ' 를 통해 신고해주세요.
오류를 발견하셨다면 해당 부분을 드래그한 후 ' 를 통해 신고해주세요.
최근 본 자료 전체보기
UCI(KEPA) : I410-ECN-0101-2020-569-000091801