이미지 캡셔닝 연구를 위해 3가지 핵심 접근 방식(Vision Encoder 기반, LLM 결합, 사전 학습 VLM)을 구현하고 비교 분석한 프로젝트입니다. 최종 모델인 ViT-Base + GPT-2는 시각-언어 특징 정렬을 통해 BLEU-4 0.1930, CIDEr-D 1.2176의 우수한 성능을 기록했습니다. 멀티모달 정렬 및 LoRA 기반의 효율적인 미세조정 실험 내용을 포함하고 있습니다.
active 2026-03-03 → 2026-03-03 (UTC)
Activity over time
Daily event counts in the loaded window
Line chart, 1 days from 2026-03-03 to 2026-03-03. Pushes: 7 total, peak 7 in a day. Pull requests: 0 total, peak 0 in a day. Issues: 0 total, peak 0 in a day. Comments: 0 total, peak 0 in a day. Stars: 0 total, peak 0 in a day.
- Pushes
- Pull requests
- Issues
- Comments
- Stars
Stars, PRs, issues and forks are under-captured in the later part of this window. GH Archive progressively stopped capturing non-push events during 2026 — −95% or worse by the end of the window. Every series here except Pushes fades for that reason, so a decline above reflects the archive, not this repository. Pushes stay reliable throughout, so read them, and the contributor counts derived from them, as the real signal. Data health has the measurements.
Top contributors
Pushes, PRs, issues, reviews and comments — stars and forks excluded, so this is contribution rather than popularity
| Contributor | Contributions | Pushes | PRs | Comments |
|---|---|---|---|---|
| SongJunPyo | 7 | 7 | 0 | 0 |
Recent activity
Latest issues, pull requests and releases
No issue or PR events — this repo's activity is pushes only.
Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 0 stars here means stars gained during the window, not the repo's star count.