권호기사보기
| 기사명 | 저자명 | 페이지 | 원문 | 기사목차 |
|---|
결과 내 검색
동의어 포함
Title Page 2
Contents 5
Abstract 6
1. Introduction 9
2. Related Work 11
3. Problem Definition 13
4. Dataset Collection 14
4.1. Processing Weather Notice 14
4.2. Processing Satellite Imagery 16
4.3. Analysis 16
4.4. Meteorological Phenomena Labels 18
5. Method 21
6. Experiments 22
6.1. Experimental Settings 22
6.2. Metric 23
6.3. Distribution Analysis 24
6.4. Verification of Weather Information Inclusion in Similar Images 26
6.5. Weather Prediction Evaluation 27
6.6. Zero-shot Weather Phenomena Search 27
6.7. Ablation Study 29
6.7.1. Evaluation of Date Relevance 29
6.7.2. Impact of Image Reconstruction Loss on Embedding Quality 30
7. Conclusion 32
References 33
국문 요약 35
Figure 1. COSMO framework 10
Figure 2. The original document of the weather notice provided by the KMA 15
Figure 3. Visualization of norms in COMS satellite images 17
Figure 4. Processed weather notice text 20
Figure 5. Visualization of embedding distributions before and after training 25
Figure 6. Similarity matrix after training (up) and before training (down) 26
Figure 7. Image-image retrieval (BBOW) 26
Figure 8. Image-image retrieval (SBERT) 26
Figure 9. Weather prediction top-k similarity (BBOW) 27
Figure 10. Weather prediction top-k similarity (SBERT) 27
Figure 11. Visualization of word embedding distribution 28
Figure 12. Comparison of retrieval results for query images 31
*표시는 필수 입력사항입니다.
| 전화번호 |
|---|
| 기사명 | 저자명 | 페이지 | 원문 | 기사목차 |
|---|
| 번호 | 발행일자 | 권호명 | 제본정보 | 자료실 | 원문 | 신청 페이지 |
|---|
도서위치안내: / 서가번호:
우편복사 목록담기를 완료하였습니다.
*표시는 필수 입력사항입니다.
저장 되었습니다.