일 | 월 | 화 | 수 | 목 | 금 | 토 |
---|---|---|---|---|---|---|
1 | 2 | 3 | 4 | |||
5 | 6 | 7 | 8 | 9 | 10 | 11 |
12 | 13 | 14 | 15 | 16 | 17 | 18 |
19 | 20 | 21 | 22 | 23 | 24 | 25 |
26 | 27 | 28 | 29 | 30 | 31 |
- gsoc가 뭔가요
- gsoc 2025
- 1차 미분 마스크
- gsoc
- 논문 리뷰
- mobilenetv1
- vlm
- grefcoco
- referring expression segmentation
- 원격 학습 안끊기게
- gres
- blip-2
- Object detection article
- 객체 검출
- 에지 검출
- 기계학습
- res
- res paper
- grefcoco dataset
- gsoc 지원
- clip
- 엔트로피란
- 이미지 필터링
- google summer of code
- 딥러닝 목적함수
- object detection
- 논문 요약
- gsoc 후기
- clip adapter
- 딥러닝 엔트로피
- Today
- Total
목록전체 글 (85)
My Vision, Computer Vision

이 글은『정성환, 배종욱, OpenCV-Python으로 배우는 영상 처리 및 응용, 생능출판2020년)』에서 공부한 내용을 토대로 작성되었다.회선(Convolution)마스크의 크기가 3 * 3이고 입력 이미지와 출력 이미지의 크기가 같다고 하자. 이 때 한 픽셀에서 컨볼루션 연산을 수행하면 그 픽셀의 주변 8픽셀까지 출력 픽셀에 관여하게 된다. 마스크의 값을 잘 조절하면 블러링, 샤프닝 등과 같이 원하는 특징을 부각시켜 출력 이미지를 만들 수 있게 된다. 블러링# 이미지 파일 경로image_path = "~~/~~/001.jpg"image = cv2.imread(image_path)image_rgb = cv2.cvtColor(image, cv2.COLOR_BGR2RGB)# 가우시안 블러링(커널 크기, ..

이 글은『크리스토퍼 비숍, 패턴 인식과 머신 러닝, 김형진 옮김, 제이펍(2018년)』에서 공부한 내용을 토대로 작성되었다.정보 이론(Information Theory)정보 이론이란 정보를 수학적으로 정의하고 측정하는 방법을 연구하는 학문이다. 머신 러닝은 데이터 정보를 다루기 때문에 정보 이론의 개념이 중요하다. (실제로 loss 함수에서 엔트로피 개념이 쓰임) 정보량(Information Content)이산 확률 변수 $x$가 있다고 하자. 이 때, 이 $x$가 가지고 있는 정보량이라는게 있다. 정보량은 ‘놀라움의 정도’라고도 하는데, 일어날 가능성이 낮은 사건이 발생했을 때의 정보량이 더 크기(더 놀랍기) 때문이다. 즉 정보량은 $x$가 발생할 확률에 반비례하며 따라서 확률 분포 $p(x)$에 종속..

Youtube 영상 장면 검색(Colab)https://github.com/gyuilLim/youtube-scene-search-with-text GitHub - gyuilLim/youtube-scene-search-with-text: Finding scenes that you want by text automaticallyFinding scenes that you want by text automatically - gyuilLim/youtube-scene-search-with-textgithub.com 장면 검색 비디오 플레이어https://github.com/gyuilLim/Scene-search-video-player GitHub - gyuilLim/Scene-search-video-player: ..

VGA: Vision GUI Assistant -- Minimizing Hallucinations through Image-Centric Fine-TuningRecent advances in Large Vision-Language Models (LVLMs) have significantly improve performance in image comprehension tasks, such as formatted charts and rich-content images. Yet, Graphical User Interface (GUI) pose a greater challenge due to their structuarxiv.orgAbstract기존 VLM은 시각적 입력을 무시하고 텍스트에 과도하게 의존하는 경..

REDQT: a method for automated mobile application GUI testing based on deep reinforcement learning algorithmsAs mobile applications become increasingly prevalent in daily life, the demand for their functionality and reliability continues to grow. Traditional mobile application testing methods, particularly graphical user interface (GUI) testing, face …www.springerprofessional.deAbstract이 논문은 심층 강..

End-to-End Object Detection with TransformersWe present a new method that views object detection as a direct set prediction problem. Our approach streamlines the detection pipeline, effectively removing the need for many hand-designed components like a non-maximum suppression procedure or anchor genearxiv.orgAbstractDETR은 Object detection을 Direct set prediction problem으로 본다.또한 NMS, Anchor genera..