DC Field | Value | Language |
---|---|---|
dc.contributor.advisor | Choo, Jaegul | - |
dc.contributor.advisor | 주재걸 | - |
dc.contributor.author | Jung, Sanghun | - |
dc.date.accessioned | 2023-06-22T19:31:23Z | - |
dc.date.available | 2023-06-22T19:31:23Z | - |
dc.date.issued | 2022 | - |
dc.identifier.uri | http://library.kaist.ac.kr/search/detail/view.do?bibCtrlNo=1008216&flag=dissertation | en_US |
dc.identifier.uri | http://hdl.handle.net/10203/308213 | - |
dc.description | 학위논문(석사) - 한국과학기술원 : 김재철AI대학원, 2022.8,[v, 24 p. :] | - |
dc.description.abstract | Identifying unexpected objects on roads in semantic segmentation (e.g., identifying dogs on roads) is crucial in safety-critical applications. Existing approaches use images of unexpected objects from external datasets or require additional training (e.g., retraining segmentation networks or training an extra network), which necessitate a non-trivial amount of labor intensity or lengthy inference time. One possible alternative is to use prediction scores of a pre-trained network such as the max logits (i.e., maximum values among classes before the final softmax layer) for detecting such objects. However, the distribution of max logits of each predicted class is significantly different from each other, which degrades the performance of identifying unexpected objects in urban-scene segmentation. To address this issue, we propose a simple yet effective approach that $\textbf{standardizes}$ the max logits in order to align the different distributions and reflect the relative meanings of max logits within each predicted class. Moreover, we consider the local regions from two different perspectives based on the intuition that neighboring pixels share similar semantic information. In contrast to previous approaches, our method does not utilize any external datasets or require additional training, which makes our method widely applicable to existing pre-trained segmentation models. Such a straightforward approach achieves a new state-of-the-art performance on the publicly available Fishyscapes Lost & Found leaderboard with a large margin. | - |
dc.language | eng | - |
dc.publisher | 한국과학기술원 | - |
dc.subject | Computer Vision▼aSemantic Segmentation▼aAutonomous Driving▼aAnomaly Detection | - |
dc.subject | 이미지 처리▼a이미지 의미분할▼a자율주행▼a이상탐지 | - |
dc.title | Standardized max logits | - |
dc.title.alternative | 도로 주행 중 이상 물체 탐지를 위한 간단하고 효과적인 방법론에 관한 연구 | - |
dc.type | Thesis(Master) | - |
dc.identifier.CNRN | 325007 | - |
dc.description.department | 한국과학기술원 :김재철AI대학원, | - |
dc.contributor.alternativeauthor | 정상헌 | - |
dc.title.subtitle | a simple yet effective approach for identifying unexpected road obstacles in urban-scene segmentation | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.