<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>Kkubuck — Computer Vision Research Notes</title><description>Paper reviews, implementation notes, and research logs on segmentation, remote sensing, and open-vocabulary learning.</description><link>https://kkubuck.github.io/</link><language>en-us</language><item><title>Seeing the Unseen: A Semantic Alignment and Context-Aware Prompt Framework for Open-Vocabulary Camouflaged Object Segmentation</title><link>https://kkubuck.github.io/papers/seeing-the-unseen-suclip/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/seeing-the-unseen-suclip/</guid><description>SuCLIP은 context-aware prompt, class-aware feature selection, semantic consistency loss, text query decoder로 OVCOS의 semantic confusion을 직접 다룬다.</description><pubDate>Sat, 04 Apr 2026 00:00:00 GMT</pubDate><category>paper</category><category>open-vocabulary</category><category>cod</category><category>clip</category><category>iccv-2025</category><author>Jisang Lee</author></item><item><title>논문 리뷰 툴을 블로그에 붙인 방법</title><link>https://kkubuck.github.io/notes/paper-review-toolkit-build-notes/</link><guid isPermaLink="true">https://kkubuck.github.io/notes/paper-review-toolkit-build-notes/</guid><description>논문 PDF를 받아 한국어 블로그 글로 정리하고 GitHub 블로그까지 반영하는 도구를, Codex가 다루기 쉬운 구조로 정리한 과정을 적었습니다.</description><pubDate>Fri, 03 Apr 2026 23:30:00 GMT</pubDate><category>codex</category><category>tooling</category><category>jekyll</category><category>automation</category><author>Jisang Lee</author></item><item><title>Rethinking Detecting Salient and Camouflaged Objects in Unconstrained Scenes</title><link>https://kkubuck.github.io/papers/uscnet-unconstrained-scenes/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/uscnet-unconstrained-scenes/</guid><description>USC12K는 두 객체 유형의 네 논리적 존재 시나리오를 주석한다. USCNet은 inter-sample·intra-sample prompt query와 CSCS metric으로 상호 혼동을 모델링·평가한다.</description><pubDate>Fri, 03 Apr 2026 00:00:00 GMT</pubDate><category>paper</category><category>sod</category><category>cod</category><category>unconstrained-scenes</category><category>iccv-2025</category><author>Jisang Lee</author></item><item><title>Beyond Single Images: Retrieval Self-Augmented Unsupervised Camouflaged Object Detection</title><link>https://kkubuck.github.io/papers/rise-unsupervised-cod/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/rise-unsupervised-cod/</guid><description>RISE는 annotation 없이 dataset-level prototype library를 만들고, Clustering-then-Retrieval과 Multi-View KNN Retrieval로 더 안정적인 pseudo mask를 생성한다.</description><pubDate>Thu, 02 Apr 2026 00:00:00 GMT</pubDate><category>paper</category><category>unsupervised-cod</category><category>retrieval</category><category>prototype-learning</category><category>iccv-2025</category><author>Jisang Lee</author></item><item><title>Improving SAM for Camouflaged Object Detection via Dual Stream Adapters</title><link>https://kkubuck.github.io/papers/sam-dsa-rgbd-cod/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/sam-dsa-rgbd-cod/</guid><description>SAM-DSA는 RGB-D dual stream adapter, model/modal bidirectional knowledge distillation, prompt update와 dual mask prediction으로 SAM을 COD에 맞춘다.</description><pubDate>Wed, 01 Apr 2026 00:00:00 GMT</pubDate><category>paper</category><category>rgb-d-cod</category><category>sam</category><category>adapter</category><category>iccv-2025</category><author>Jisang Lee</author></item><item><title>Enhancing Prompt Generation with Adaptive Refinement for Camouflaged Object Detection</title><link>https://kkubuck.github.io/papers/adaptive-refinement-arm-cod/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/adaptive-refinement-arm-cod/</guid><description>ARM은 multimodal 정보와 mask prompt를 동시에 보정하고, refinement 과정의 중간 표현을 auxiliary embedding으로 재사용해 SAM에 더 풍부한 조건을 제공한다.</description><pubDate>Tue, 31 Mar 2026 00:00:00 GMT</pubDate><category>paper</category><category>sam</category><category>prompt-refinement</category><category>multimodal</category><category>iccv-2025</category><author>Jisang Lee</author></item><item><title>ESCNet: Edge-Semantic Collaborative Network for Camouflaged Object Detection</title><link>https://kkubuck.github.io/papers/escnet-edge-semantic-cod/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/escnet-edge-semantic-cod/</guid><description>AETP가 edge–texture를 공동 인식하고, DSFA가 local complexity와 orientation에 맞춰 sampling하며, MFMM이 계층 특징을 점진적으로 보정한다.</description><pubDate>Mon, 30 Mar 2026 00:00:00 GMT</pubDate><category>paper</category><category>cod</category><category>edge-texture</category><category>dynamic-sampling</category><category>iccv-2025</category><author>Jisang Lee</author></item><item><title>Multi-modal Segment Anything Model for Camouflaged Scene Segmentation</title><link>https://kkubuck.github.io/papers/mm-sam-camouflaged-scene-segmentation/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/mm-sam-camouflaged-scene-segmentation/</guid><description>BLIP text·visual embedding, multi-level adapter, SAM dense embedding 교체를 통해 prompt-free multimodal COD를 구현하고 다른 segmentation task로도 확장한다.</description><pubDate>Sun, 29 Mar 2026 00:00:00 GMT</pubDate><category>paper</category><category>sam</category><category>vision-language</category><category>multimodal</category><category>iccv-2025</category><author>Jisang Lee</author></item><item><title>Scoring, Remember, and Reference: Catching Camouflaged Objects in Videos</title><link>https://kkubuck.github.io/papers/srrnet-video-camouflaged-objects/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/srrnet-video-camouflaged-objects/</guid><description>SRR은 dual-purpose decoder로 mask와 reference score를 생성하고, reference-guided multilevel asymmetric attention으로 memory와 motion을 통합한다.</description><pubDate>Sat, 28 Mar 2026 00:00:00 GMT</pubDate><category>paper</category><category>video-cod</category><category>memory</category><category>reference-frame</category><category>iccv-2025</category><author>Jisang Lee</author></item><item><title>Shift the Lens: Environment-Aware Unsupervised Camouflaged Object Detection</title><link>https://kkubuck.github.io/papers/ease-environment-aware-unsupervised-cod/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/ease-environment-aware-unsupervised-cod/</guid><description>EASE는 multimodal·diffusion·vision foundation model로 environment prototype library를 만들고, KDE-AT·G2L·Self-Retrieval로 장면에 맞는 배경을 찾는다.</description><pubDate>Fri, 27 Mar 2026 00:00:00 GMT</pubDate><category>paper</category><category>unsupervised-cod</category><category>retrieval</category><category>environment-modeling</category><category>cvpr-2025</category><author>Jisang Lee</author></item><item><title>UCOD-DPL: Unsupervised Camouflaged Object Detection via Dynamic Pseudo-label Learning</title><link>https://kkubuck.github.io/papers/ucod-dpl-dynamic-pseudo-label-learning/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/ucod-dpl-dynamic-pseudo-label-learning/</guid><description>UCOD-DPL은 teacher–student 구조 안에서 Adaptive Pseudo-label Module, Dual-Branch Adversarial decoder, Look-Twice refinement를 결합한다.</description><pubDate>Thu, 26 Mar 2026 00:00:00 GMT</pubDate><category>paper</category><category>unsupervised-cod</category><category>pseudo-label</category><category>teacher-student</category><category>cvpr-2025</category><author>Jisang Lee</author></item><item><title>COD 벤치마크를 읽는 법: 데이터셋, 지표, 비교 체크리스트</title><link>https://kkubuck.github.io/papers/cod-benchmarks-guide/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/cod-benchmarks-guide/</guid><description>COD 논문의 표를 그대로 비교하기 전에 데이터 분할, backbone, 입력 크기, 추가 supervision과 metric의 성격을 확인하는 실전 가이드.</description><pubDate>Wed, 25 Mar 2026 00:00:00 GMT</pubDate><category>guide</category><category>cod</category><category>benchmark</category><category>dataset</category><category>evaluation</category><author>Jisang Lee</author></item><item><title>Open-Vocabulary Camouflaged Object Segmentation</title><link>https://kkubuck.github.io/papers/ovcos-ovcoser-eccv2024/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/ovcos-ovcoser-eccv2024/</guid><description>OVCOS라는 새 task와 OVCamo dataset을 제안한다. OVCoser는 고정된 CLIP에 iterative semantic guidance와 edge/depth structure enhancement를 붙여 novel class를 분할한다.</description><pubDate>Tue, 24 Mar 2026 00:00:00 GMT</pubDate><category>paper</category><category>open-vocabulary</category><category>cod</category><category>vision-language</category><category>eccv-2024</category><author>Jisang Lee</author></item><item><title>Learning Camouflaged Object Detection from Noisy Pseudo Label</title><link>https://kkubuck.github.io/papers/noisy-pseudo-label-cod-eccv2024/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/noisy-pseudo-label-cod-eccv2024/</guid><description>제한된 fully labeled data와 box-prompt pseudo labels를 결합하고, noise correction loss가 학습 단계에 따라 올바른 픽셀을 우선 학습하고 noisy gradient를 수정한다.</description><pubDate>Mon, 23 Mar 2026 00:00:00 GMT</pubDate><category>paper</category><category>weakly-semi-supervised</category><category>pseudo-label</category><category>label-noise</category><category>eccv-2024</category><author>Jisang Lee</author></item><item><title>Just a Hint: Point-Supervised Camouflaged Object Detection</title><link>https://kkubuck.github.io/papers/just-a-hint-point-supervised-cod-eccv2024/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/just-a-hint-point-supervised-cod-eccv2024/</guid><description>한 점 supervision에서 adaptive hint area를 만들고 attention regulator와 unsupervised contrastive learning으로 partial localization과 불안정한 특징을 보완한다.</description><pubDate>Sun, 22 Mar 2026 00:00:00 GMT</pubDate><category>paper</category><category>weak-supervision</category><category>point-supervision</category><category>cod</category><category>eccv-2024</category><author>Jisang Lee</author></item><item><title>Frequency-Spatial Entanglement Learning for Camouflaged Object Detection</title><link>https://kkubuck.github.io/papers/fsel-cod-eccv2024/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/fsel-cod-eccv2024/</guid><description>FSEL은 ETB의 Frequency Self-Attention과 Entanglement FFN, Joint Domain Perception, Dual-domain Reverse Parser로 두 도메인을 깊게 상호작용시킨다.</description><pubDate>Sat, 21 Mar 2026 00:00:00 GMT</pubDate><category>paper</category><category>cod</category><category>frequency-spatial</category><category>transformer</category><category>eccv-2024</category><author>Jisang Lee</author></item><item><title>VSCode: General Visual Salient and Camouflaged Object Detection with 2D Prompt Learning</title><link>https://kkubuck.github.io/papers/vscode-generalist-cod-cvpr2024/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/vscode-generalist-cod-cvpr2024/</guid><description>VSCode는 VST 기반 encoder–decoder에 domain prompt와 task prompt를 분리해 넣는다. 네 SOD와 세 COD 작업을 공동 학습하고 보지 못한 조합으로 전이한다.</description><pubDate>Fri, 20 Mar 2026 00:00:00 GMT</pubDate><category>paper</category><category>generalist-model</category><category>prompt-learning</category><category>cod</category><category>sod</category><category>cvpr-2024</category><author>Jisang Lee</author></item><item><title>Endow SAM with Keen Eyes: Temporal-spatial Prompt Learning for Video Camouflaged Object Detection</title><link>https://kkubuck.github.io/papers/tsp-sam-vcod-cvpr2024/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/tsp-sam-vcod-cvpr2024/</guid><description>TSP-SAM은 motion-driven self-prompt와 long-range consistency를 결합한다. temporal cue를 prompt 생성과 SAM encoder adaptation 양쪽에 사용한다.</description><pubDate>Thu, 19 Mar 2026 00:00:00 GMT</pubDate><category>paper</category><category>video-cod</category><category>sam</category><category>prompt-learning</category><category>cvpr-2024</category><author>Jisang Lee</author></item><item><title>The Making and Breaking of Camouflage</title><link>https://kkubuck.github.io/papers/making-and-breaking-of-camouflage-iccv2023/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/making-and-breaking-of-camouflage-iccv2023/</guid><description>foreground–background feature similarity와 boundary visibility로 camouflage effectiveness를 측정하고, 이를 생성 모델의 auxiliary loss로 사용해 VCOD 학습 데이터를 만든다.</description><pubDate>Wed, 18 Mar 2026 00:00:00 GMT</pubDate><category>paper</category><category>camouflage-analysis</category><category>synthetic-data</category><category>video-cod</category><category>iccv-2023</category><author>Jisang Lee</author></item><item><title>Source-free Depth for Object Pop-out</title><link>https://kkubuck.github.io/papers/source-free-depth-pop-out-iccv2023/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/source-free-depth-pop-out-iccv2023/</guid><description>PopNet은 pretrained monocular depth model을 source-free로 적응시킨다. contact surface를 약한 mask supervision으로 학습해 3D에서 객체와 배경을 분리한다.</description><pubDate>Tue, 17 Mar 2026 00:00:00 GMT</pubDate><category>paper</category><category>depth</category><category>source-free-adaptation</category><category>cod</category><category>sod</category><category>iccv-2023</category><author>Jisang Lee</author></item><item><title>Camouflaged Object Detection with Feature Decomposition and Edge Reconstruction</title><link>https://kkubuck.github.io/papers/feder-cod-cvpr2023/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/feder-cod-cvpr2023/</guid><description>FEDER는 foreground–background 유사성은 frequency decomposition으로, ambiguous boundary는 auxiliary edge reconstruction으로 분리해 해결한다.</description><pubDate>Mon, 16 Mar 2026 00:00:00 GMT</pubDate><category>paper</category><category>cod</category><category>wavelet</category><category>edge-reconstruction</category><category>cvpr-2023</category><author>Jisang Lee</author></item><item><title>Feature Shrinkage Pyramid for Camouflaged Object Detection with Transformers</title><link>https://kkubuck.github.io/papers/fspnet-transformer-cod-cvpr2023/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/fspnet-transformer-cod-cvpr2023/</guid><description>FSPNet은 NL-TEM으로 인접 token의 고차 관계를 강화하고, AIM을 포함한 Feature Shrinkage Decoder가 이웃 transformer feature를 층별로 모은다.</description><pubDate>Sun, 15 Mar 2026 00:00:00 GMT</pubDate><category>paper</category><category>cod</category><category>transformer</category><category>feature-pyramid</category><category>cvpr-2023</category><author>Jisang Lee</author></item><item><title>Camouflaged Instance Segmentation via Explicit De-Camouflaging</title><link>https://kkubuck.github.io/papers/dcnet-cis-cvpr2023/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/dcnet-cis-cvpr2023/</guid><description>DCNet은 pixel-level camouflage decoupling과 instance-level camouflage suppression을 결합한다. semantic mask가 아니라 여러 위장 개체를 분리해야 하는 CIS에 맞춘 설계다.</description><pubDate>Sat, 14 Mar 2026 00:00:00 GMT</pubDate><category>paper</category><category>camouflaged-instance-segmentation</category><category>fourier</category><category>instance-prototype</category><category>cvpr-2023</category><author>Jisang Lee</author></item><item><title>Detecting Camouflaged Object in Frequency Domain</title><link>https://kkubuck.github.io/papers/fdcod-cvpr2022/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/fdcod-cvpr2022/</guid><description>FDCOD은 DCT 기반 Frequency Enhancement Module, RGB–frequency feature alignment, High-Order Relation Module을 결합해 공간 영역 밖의 단서를 적극적으로 사용한다.</description><pubDate>Fri, 13 Mar 2026 00:00:00 GMT</pubDate><category>paper</category><category>cod</category><category>frequency-domain</category><category>feature-fusion</category><category>cvpr-2022</category><author>Jisang Lee</author></item><item><title>Zoom in and Out: A Mixed-Scale Triplet Network for Camouflaged Object Detection</title><link>https://kkubuck.github.io/papers/zoomnet-cvpr2022/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/zoomnet-cvpr2022/</guid><description>ZoomNet은 mixed-scale triplet, Scale Integration Unit, Hierarchical Mixed-scale Unit으로 확대·축소 관찰을 모사한다. uncertainty-aware loss로 애매한 후보 영역의 confidence도 다룬다.</description><pubDate>Thu, 12 Mar 2026 00:00:00 GMT</pubDate><category>paper</category><category>cod</category><category>multi-scale</category><category>uncertainty</category><category>cvpr-2022</category><author>Jisang Lee</author></item><item><title>Segment, Magnify and Reiterate: Detecting Camouflaged Objects the Hard Way</title><link>https://kkubuck.github.io/papers/segmar-cvpr2022/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/segmar-cvpr2022/</guid><description>SegMaR은 Segment–Magnify–Reiterate의 반복 구조를 사용한다. discriminative mask와 attention sampler로 작은 객체와 저해상도 경계를 선택적으로 확대한다.</description><pubDate>Wed, 11 Mar 2026 00:00:00 GMT</pubDate><category>paper</category><category>cod</category><category>iterative-refinement</category><category>small-objects</category><category>cvpr-2022</category><author>Jisang Lee</author></item><item><title>Implicit Motion Handling for Video Camouflaged Object Detection</title><link>https://kkubuck.github.io/papers/implicit-motion-vcod-cvpr2022/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/implicit-motion-vcod-cvpr2022/</guid><description>인접 프레임의 dense correlation volume으로 단기 움직임을 포착하고 spatio-temporal transformer로 장기 일관성을 보강한다. MoCA-Mask 데이터셋도 함께 제안한다.</description><pubDate>Tue, 10 Mar 2026 00:00:00 GMT</pubDate><category>paper</category><category>video-cod</category><category>motion</category><category>temporal-consistency</category><category>cvpr-2022</category><author>Jisang Lee</author></item><item><title>Uncertainty-Guided Transformer Reasoning for Camouflaged Object Detection</title><link>https://kkubuck.github.io/papers/ugtr-cod-iccv2021/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/ugtr-cod-iccv2021/</guid><description>UGTR은 backbone 출력의 조건부 분포에서 초기 예측과 uncertainty를 얻고, 불확실한 영역을 attention으로 재추론한다. Bayesian representation과 transformer를 결합한 초기 COD 연구다.</description><pubDate>Mon, 09 Mar 2026 00:00:00 GMT</pubDate><category>paper</category><category>cod</category><category>uncertainty</category><category>transformer</category><category>iccv-2021</category><author>Jisang Lee</author></item><item><title>Camouflaged Object Segmentation with Distraction Mining</title><link>https://kkubuck.github.io/papers/pfnet-distraction-mining-cvpr2021/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/pfnet-distraction-mining-cvpr2021/</guid><description>PFNet은 포식 과정에서 영감을 얻어 전역 위치 탐색과 국소 식별을 분리한다. Focus Module은 모호한 영역의 distraction을 발견하고 제거하며 거친 예측을 단계적으로 다듬는다.</description><pubDate>Sun, 08 Mar 2026 00:00:00 GMT</pubDate><category>paper</category><category>cod</category><category>coarse-to-fine</category><category>distraction-mining</category><category>cvpr-2021</category><author>Jisang Lee</author></item><item><title>Uncertainty-Aware Joint Salient Object and Camouflaged Object Detection</title><link>https://kkubuck.github.io/papers/joint-sod-cod-cvpr2021/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/joint-sod-cod-cvpr2021/</guid><description>SOD와 COD의 상반된 속성을 활용해 두 작업을 함께 개선한다. 쉬운 COD 샘플을 SOD의 어려운 양성으로 사용하고, 유사도 측정과 adversarial uncertainty 학습을 결합한다.</description><pubDate>Sat, 07 Mar 2026 00:00:00 GMT</pubDate><category>paper</category><category>cod</category><category>sod</category><category>uncertainty</category><category>multi-task</category><category>cvpr-2021</category><author>Jisang Lee</author></item><item><title>Simultaneously Localize, Segment and Rank the Camouflaged Objects</title><link>https://kkubuck.github.io/papers/rank-camouflaged-objects-cvpr2021/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/rank-camouflaged-objects-cvpr2021/</guid><description>이 논문은 객체 위치·분할·위장 난도 순위를 동시에 학습한다. 이진 마스크만으로는 표현할 수 없던 “얼마나 잘 숨었는가”를 COD의 명시적 출력으로 만든다.</description><pubDate>Fri, 06 Mar 2026 00:00:00 GMT</pubDate><category>paper</category><category>cod</category><category>ranking</category><category>multi-task</category><category>cvpr-2021</category><author>Jisang Lee</author></item><item><title>Mutual Graph Learning for Camouflaged Object Detection</title><link>https://kkubuck.github.io/papers/mutual-graph-learning-cod-cvpr2021/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/mutual-graph-learning-cod-cvpr2021/</guid><description>MGL은 위치와 경계를 별도 그래프로 표현한 뒤 상호 보완 관계를 반복적으로 학습한다. 픽셀 자체보다 객체·경계 사이의 관계를 COD의 핵심 단서로 본 작업이다.</description><pubDate>Thu, 05 Mar 2026 00:00:00 GMT</pubDate><category>paper</category><category>cod</category><category>graph-reasoning</category><category>cvpr-2021</category><author>Jisang Lee</author></item><item><title>Unsupervised Domain Adaptation for SAR Target Classification Based on Domain- and Class-level Alignment</title><link>https://kkubuck.github.io/papers/tistory-44/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/tistory-44/</guid><description>시뮬레이션→실측 SAR 전이에서 전역 분포뿐 아니라 클래스 원형과 물리 단서까지 이용해 잘못된 정렬을 줄인다.</description><pubDate>Thu, 22 Feb 2024 08:34:52 GMT</pubDate><category>paper</category><category>domain-adaptation</category><category>sar</category><category>classification</category><category>contrastive-learning</category><author>Jisang Lee</author></item><item><title>Unsupervised Domain Adaptation Based on Progressive Transfer for Ship Detection: From Optical to SAR Images</title><link>https://kkubuck.github.io/papers/tistory-43/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/tistory-43/</guid><description>광학→SAR의 큰 간극을 픽셀·특징·예측 단계로 순차 축소하고, 마지막에는 robust self-training으로 target 표현을 직접 학습한다.</description><pubDate>Mon, 19 Feb 2024 10:07:03 GMT</pubDate><category>paper</category><category>domain-adaptation</category><category>sar</category><category>ship-detection</category><category>self-training</category><author>Jisang Lee</author></item><item><title>Unsupervised Domain-Adaptive SAR Ship Detection Based on Cross-Domain Feature Interaction and Data Contribution Balance</title><link>https://kkubuck.github.io/papers/tistory-42/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/tistory-42/</guid><description>광학과 SAR의 큰 분포 차이를 영상 생성·특징 추출·작은 선박 강화·샘플 가중의 네 단계로 줄이는 도메인 적응 방법.</description><pubDate>Mon, 19 Feb 2024 10:05:32 GMT</pubDate><category>paper</category><category>domain-adaptation</category><category>sar</category><category>object-detection</category><category>remote-sensing</category><author>Jisang Lee</author></item><item><title>인공지능 텀프로젝트(2023-2)</title><link>https://kkubuck.github.io/notes/tistory-41/</link><guid isPermaLink="true">https://kkubuck.github.io/notes/tistory-41/</guid><description>인공지능 텀프로젝트(2023-2) 자료를 기존 Tistory 블로그에서 이전했습니다.</description><pubDate>Thu, 15 Feb 2024 04:29:44 GMT</pubDate><category>Major class</category><category>인공지능</category><author>Jisang Lee</author></item><item><title>Open-Vocabulary Semantic Segmentation with Mask-adapted CLIP</title><link>https://kkubuck.github.io/papers/tistory-40/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/tistory-40/</guid><description>OVSeg는 마스크 영역을 분류할 때 CLIP이 급격히 약해지는 병목을 찾아, 데이터와 프롬프트 양쪽에서 입력 분포를 맞춘다.</description><pubDate>Thu, 15 Feb 2024 04:25:06 GMT</pubDate><category>paper</category><category>open-vocabulary</category><category>semantic-segmentation</category><category>clip</category><author>Jisang Lee</author></item><item><title>Activating More Pixels in Image Super-Resolution Transformer</title><link>https://kkubuck.github.io/papers/tistory-39/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/tistory-39/</guid><description>HAT은 창 기반 Transformer가 실제로 참고하는 입력 범위가 좁다는 관찰에서 출발해 채널·공간 주의를 함께 확장한다.</description><pubDate>Thu, 16 Nov 2023 13:32:29 GMT</pubDate><category>paper</category><category>super-resolution</category><category>transformer</category><category>image-restoration</category><author>Jisang Lee</author></item><item><title>FeNet: Feature Enhancement Network for Lightweight Remote-Sensing Image Super-Resolution</title><link>https://kkubuck.github.io/papers/tistory-38/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/tistory-38/</guid><description>원격탐사 초해상도에서 작은 모델이 세부 질감을 놓치지 않도록 격자형 특징 교환과 계층적 강화를 설계한 네트워크.</description><pubDate>Thu, 16 Nov 2023 13:31:11 GMT</pubDate><category>paper</category><category>super-resolution</category><category>remote-sensing</category><category>lightweight</category><author>Jisang Lee</author></item><item><title>From Beginner to Master: A Survey for Deep Learning-based Single-Image Super-Resolution</title><link>https://kkubuck.github.io/papers/tistory-37/</link><guid isPermaLink="true">https://kkubuck.github.io/papers/tistory-37/</guid><description>SISR 논문을 모델 이름이 아니라 목표·데이터·손실·평가의 축으로 읽게 해 주는 입문용 지도.</description><pubDate>Thu, 16 Nov 2023 13:26:08 GMT</pubDate><category>paper</category><category>super-resolution</category><category>survey</category><category>image-restoration</category><author>Jisang Lee</author></item><item><title>퍼셉트론, 신경망</title><link>https://kkubuck.github.io/notes/tistory-36/</link><guid isPermaLink="true">https://kkubuck.github.io/notes/tistory-36/</guid><description>1. 퍼셉트론 1-1. 퍼셉트론이란? 다수의 신호를 입력으로 받아 하나의 신호 를 출력한다. 퍼셉트론 신호는 0/1의 두가지 값을 가질 수 있다. 입력이 2개인 퍼셉트론 그림의 원을 노드 혹은 뉴런이라고 부른다. 입력 신호가 뉴런에 보내질 때는 각각 고유한 가중치(w1, w2) 가 곱해진다. 뉴런에서 보내온 신호의 총합이</description><pubDate>Sat, 17 Dec 2022 12:02:43 GMT</pubDate><category>Lab</category><category>밑바닥부터 시작하는 딥러닝</category><category>기계학습</category><category>딥러닝</category><category>렐루</category><category>분류</category><category>순전파</category><category>시그모이드</category><category>신경망</category><category>퍼셉트론</category><category>활성화함수</category><author>Jisang Lee</author></item><item><title>2022 KAKAO TECH INTERNSHIP: 성격 유형 검사하기</title><link>https://kkubuck.github.io/notes/tistory-35/</link><guid isPermaLink="true">https://kkubuck.github.io/notes/tistory-35/</guid><description>문제 설명 나만의 카카오 성격 유형 검사지를 만들려고 합니다. 성격 유형 검사는 다음과 같은 4개 지표로 성격 유형을 구분합니다. 성격은 각 지표에서 두 유형 중 하나로 결정됩니다. 지표 번호성격 유형 1번 지표 라이언형(R), 튜브형(T) 2번 지표 콘형(C), 프로도형(F) 3번 지표 제이지형(J), 무지형(M) 4번</description><pubDate>Tue, 06 Sep 2022 04:26:58 GMT</pubDate><category>Coding test</category><category>프로그래머스</category><category>백준</category><category>알고리즘</category><category>코딩테스트</category><category>파이썬</category><author>Jisang Lee</author></item><item><title>백준 11650번 Python</title><link>https://kkubuck.github.io/notes/tistory-34/</link><guid isPermaLink="true">https://kkubuck.github.io/notes/tistory-34/</guid><description>n = int(input()) x = [] for i in range(n): a, b = map(int, input().split()) x.append((a, b)) x.sort() for i in range(n): print(x[i][0], x[i][1])</description><pubDate>Mon, 05 Sep 2022 02:17:24 GMT</pubDate><category>Coding test</category><category>백준</category><category>11650</category><category>파이썬</category><author>Jisang Lee</author></item><item><title>백준 14681번 C++</title><link>https://kkubuck.github.io/notes/tistory-33/</link><guid isPermaLink="true">https://kkubuck.github.io/notes/tistory-33/</guid><description>#include &lt;iostream&gt; using namespace std; int main(){ int a,b; cin&gt;&gt;a; cin&gt;&gt;b; if((a&gt;=-1000&amp;&amp;a&lt;=1000)&amp;&amp;(b&gt;=-1000&amp;&amp;b&lt;=1000)){ if(a&gt;0&amp;&amp;b&gt;0){ cout&lt;&lt;&quot;1&quot;; } if(a&lt;0&amp;&amp;b&gt;0){ cout&lt;&lt;&quot;2&quot;; } if</description><pubDate>Mon, 05 Sep 2022 02:16:15 GMT</pubDate><category>Coding test</category><category>백준</category><category>14681</category><category>C++</category><author>Jisang Lee</author></item><item><title>백준 15596번 Python</title><link>https://kkubuck.github.io/notes/tistory-32/</link><guid isPermaLink="true">https://kkubuck.github.io/notes/tistory-32/</guid><description>def solve(a): ans = sum(a) return ans</description><pubDate>Mon, 05 Sep 2022 02:14:54 GMT</pubDate><category>Coding test</category><category>백준</category><category>15596</category><category>파이썬</category><category>함수</category><author>Jisang Lee</author></item><item><title>백준 18258번 Python</title><link>https://kkubuck.github.io/notes/tistory-31/</link><guid isPermaLink="true">https://kkubuck.github.io/notes/tistory-31/</guid><description>import sys from collections import deque n = int(sys.stdin.readline().rstrip()) queue =deque([]) for i in range(n): ans = sys.stdin.readline().rstrip().split() if ans[0]==&apos;push&apos;: q</description><pubDate>Mon, 05 Sep 2022 02:13:40 GMT</pubDate><category>Coding test</category><category>백준</category><category>18258</category><category>알고리즘</category><category>큐</category><author>Jisang Lee</author></item><item><title>백준 25304번 Python</title><link>https://kkubuck.github.io/notes/tistory-30/</link><guid isPermaLink="true">https://kkubuck.github.io/notes/tistory-30/</guid><description>all = int(input()) num = int(input()) for i in range(num): price, n = map(int, input().split()) all-=price * n if all == 0: print(&apos;Yes&apos;) else: print(&apos;No&apos;)</description><pubDate>Mon, 05 Sep 2022 02:12:30 GMT</pubDate><category>Coding test</category><category>백준</category><category>25304</category><category>구현</category><category>사칙연산</category><category>수학</category><category>영수증</category><author>Jisang Lee</author></item><item><title>SW중심대학 공동 AI 경진대회 &lt;예선&gt; 후기</title><link>https://kkubuck.github.io/notes/tistory-29/</link><guid isPermaLink="true">https://kkubuck.github.io/notes/tistory-29/</guid><description>8월부터 9월까지 진행됬던 SW중심대학 공동 AI 경진대회 &lt;예선&gt; 가 드디어 막을 내렸습니다. 제가 속한 팀인 AIM. Lab.은 대회 중 계산하는 스코어인 Public에서는 최종 2위 , 최종 스코어인 Private 에서는 1위 를 달성했습니다. 대회 주제는 &apos;&apos;심리학 테스트 데이터를 분석하여 &quot;심리 성향을 예측&quot;하는</description><pubDate>Mon, 05 Sep 2022 02:06:32 GMT</pubDate><category>Lab</category><category>혼자 공부하는 머신러닝 + 딥러닝</category><category>DACON</category><category>DL</category><category>ml</category><category>SW중심대학</category><category>SW중심대학 공동 AI 경진대회 &lt;예선&gt;</category><category>경진대회</category><category>데이터분석</category><category>딥러닝</category><category>머신러닝</category><category>정형데이터</category><author>Jisang Lee</author></item><item><title>[ 혼자 공부하는 머신러닝 + 딥러닝 ] 트리의 앙상블</title><link>https://kkubuck.github.io/notes/tistory-27/</link><guid isPermaLink="true">https://kkubuck.github.io/notes/tistory-27/</guid><description>가지런히 정리 되어 있는 CSV파일, 엑셀파일등을 정형 데이터 라고 부른다. 글과 같은 텍스트 데이터, 사진, 음악 등을 비정형 데이터 라고 부른다. 정형 데이터 를 다루는데 가장 뛰어난 성과 를 내는 알고리즘이 앙상블 학습 이다. 랜덤 포레스트 앙상블 학습의 대표 주자 중 하나로 안정적인 성능 을 낸다. 결정트리를 랜덤</description><pubDate>Mon, 04 Jul 2022 04:12:04 GMT</pubDate><category>Lab</category><category>혼자 공부하는 머신러닝 + 딥러닝</category><category>그레이디언트 부스팅</category><category>딥러닝</category><category>랜덤포레스트</category><category>머신러닝</category><category>부트스트랩</category><category>앙상블 학습</category><category>엑스트라트리</category><category>트리의 앙상블</category><category>혼자공부하는머신러닝+딥러닝</category><category>히스토그램 기반 그레이디언트 부스팅</category><author>Jisang Lee</author></item><item><title>[ 혼자 공부하는 머신러닝 + 딥러닝 ] 교차 검증과 그리드 서치</title><link>https://kkubuck.github.io/notes/tistory-26/</link><guid isPermaLink="true">https://kkubuck.github.io/notes/tistory-26/</guid><description>검증세트 테스트세트 를 사용하지 않고 모델이 과소적합인지 과대적합인지 판단하기 위해 훈련세트 를 또 나눠서 검증세트 를 만든다. 훈련세트 에서 모델을 훈련 하고 검증 세트 로 모델을 평가 한다. from sklearn.model_selection import train_test_split #데이터와 타깃을 훈련세트와 테스</description><pubDate>Sun, 03 Jul 2022 10:22:37 GMT</pubDate><category>Lab</category><category>혼자 공부하는 머신러닝 + 딥러닝</category><category>검증세트</category><category>교차검증</category><category>그리드서치</category><category>딥러닝</category><category>랜덤서치</category><category>머신러닝</category><category>모델파라미터</category><category>하이퍼파라미터</category><category>혼공머신</category><category>혼자공부하는머신러닝+딥러닝</category><author>Jisang Lee</author></item><item><title>[ 혼자 공부하는 머신러닝 + 딥러닝 ] 결정 트리</title><link>https://kkubuck.github.io/notes/tistory-25/</link><guid isPermaLink="true">https://kkubuck.github.io/notes/tistory-25/</guid><description>판다스 데이터프레임의 유용한 메서드 데이터프레임 객체.info () → 데이터프레임의 열의 데이터 타입과 누락된 데이터 가 있는지 확인 데이터프레임 객체.describe() → 열에 대한 간략한 통계 를 출력(mean, std, min, max 등등) 데이터프레임에서 각 통계 확인 후 특성에 대한 스케일 이 다를경우 St</description><pubDate>Sun, 03 Jul 2022 05:29:51 GMT</pubDate><category>Lab</category><category>혼자 공부하는 머신러닝 + 딥러닝</category><category>가지치기</category><category>결정트리</category><category>기계학습</category><category>딥러닝</category><category>머신러닝</category><category>불순도</category><category>엔트로피불순도</category><category>지니불순도</category><category>혼공머신</category><category>혼자공부하는머신러닝+딥러닝</category><author>Jisang Lee</author></item><item><title>[ 혼자 공부하는 머신러닝 + 딥러닝 ] 확률적 경사 하강법</title><link>https://kkubuck.github.io/notes/tistory-23/</link><guid isPermaLink="true">https://kkubuck.github.io/notes/tistory-23/</guid><description>점진적 학습 이전에 학습한 모델을 버리지 않고 새로운 데이터셋에 대해서 조금씩 훈련 대표적인 점진적 학습 알고리즘 은 확률적 경사 하강법 입니다. 확률적 경사 하강법 경사를 따라 내려가는 방법 가파른 경사를 따라 원하는 지점 에 도달하는것이 목표 가파른 길을 찾아 내려오지만 조금씩 내려오는게 중요 → 경사 하강법 모델 을</description><pubDate>Sat, 02 Jul 2022 12:24:45 GMT</pubDate><category>Lab</category><category>혼자 공부하는 머신러닝 + 딥러닝</category><category>기계학습</category><category>딥러닝</category><category>로지스틱 손실</category><category>로지스틱회귀</category><category>머신러닝</category><category>에포크</category><category>인공지능</category><category>컴퓨터공학과</category><category>혼자공부하는머신러닝+딥러닝</category><category>확률적경사하강법</category><author>Jisang Lee</author></item><item><title>[ 혼자 공부하는 머신러닝 + 딥러닝 ] 로지스틱 회귀</title><link>https://kkubuck.github.io/notes/tistory-22/</link><guid isPermaLink="true">https://kkubuck.github.io/notes/tistory-22/</guid><description>로지스틱 회귀 데이터 준비하기 CSV 파일을 읽을 경우 데이터프레임 형태로 저장된다. 데이터프레임에서 사용할 열을 입력데이터 와 타깃데이터 로 구분해 넘파이 배열 로 변환한다. 분류한 입력과 타깃을 사이킷런의 train_test_split 함수를 활용하여 훈련데이터 와 테스트데이터 로 분할한다. from sklearn.m</description><pubDate>Sat, 02 Jul 2022 06:47:57 GMT</pubDate><category>Lab</category><category>혼자 공부하는 머신러닝 + 딥러닝</category><category>로지스틱회귀</category><category>사이파이</category><category>소프트맥스</category><category>시그모이드</category><category>혼공머</category><category>혼공머신</category><category>혼자공부하는머신러닝+딥러닝</category><author>Jisang Lee</author></item><item><title>[ 혼자 공부하는 머신러닝 + 딥러닝 ] 마켓과 머신러닝</title><link>https://kkubuck.github.io/notes/tistory-21/</link><guid isPermaLink="true">https://kkubuck.github.io/notes/tistory-21/</guid><description>생선들의 데이터 분류 방법 일단 생선의 특징을 알아야 합니다. = 길이, 무게등등 찾고자 하는 생선의 데이터(길이, 무게)를 리스트로 준비합니다. 머신러닝에서 여러개의 종류(클래스) 중 하나를 구별해 내는 문제를 분류(classification) 이라고 부릅니다. 2개의 클래스중 하나를 고르는 문제는 이진 분류(binar</description><pubDate>Tue, 21 Jun 2022 10:19:35 GMT</pubDate><category>Lab</category><category>혼자 공부하는 머신러닝 + 딥러닝</category><category>컴퓨터공학과</category><category>데이터분석</category><category>딥러닝</category><category>머신러닝</category><category>사이킷런</category><category>알고리즘</category><category>모델</category><category>k-최근접 이웃 알고리즘</category><author>Jisang Lee</author></item><item><title>[ 혼자 공부하는 머신러닝 + 딥러닝 ] 인공지능과 머신러닝, 딥러닝</title><link>https://kkubuck.github.io/notes/tistory-20/</link><guid isPermaLink="true">https://kkubuck.github.io/notes/tistory-20/</guid><description>알파고가 등장하기 전 인공지능은 소설이나 영화속에서만 보여져왔습니다. 하지만 알파고의 등장 이후 현실 속의 우리들과 밀접하게 현실 속의 기술로 급격히 발전중입니다. 인공지능 은 사람처럼 학습하고 추론 가능한 지능을 가진 컴퓨터 시스템을 만드는 기술입니다. 1. 인공지능의 역사 1943년: 워런 매컬러와 윌터 피츠는 최초로</description><pubDate>Tue, 21 Jun 2022 08:09:10 GMT</pubDate><category>Lab</category><category>혼자 공부하는 머신러닝 + 딥러닝</category><category>컴퓨터공학과</category><category>데이터분석</category><category>인공지능</category><category>머신러닝</category><category>딥러닝</category><category>라이브러리</category><author>Jisang Lee</author></item><item><title>[ 자료구조 ] 동적 계획법 (dynamic programming)</title><link>https://kkubuck.github.io/notes/tistory-19/</link><guid isPermaLink="true">https://kkubuck.github.io/notes/tistory-19/</guid><description>1. 배낭에 담을 수 있는 보석의 최대 가격 계산 함수 def knapsack(maxWeight, rowCount, weight, money): &quot;&quot;&quot; maxWeight: 배낭 최대 무게 rowCount: 보석 개수 weight: 보석 무게 리스트 money: 보석 가격 리스트 출력값: 배낭에 담을 수 있는 보석의 최대</description><pubDate>Tue, 21 Jun 2022 07:00:00 GMT</pubDate><category>Major class</category><category>자료구조</category><category>컴퓨터공학과</category><category>동적계획법</category><author>Jisang Lee</author></item><item><title>[ 자료구조 ] 검색 알고리즘 (search algorithm)</title><link>https://kkubuck.github.io/notes/tistory-18/</link><guid isPermaLink="true">https://kkubuck.github.io/notes/tistory-18/</guid><description>1. 이진 검색 함수 def book_search(index_array, find_name): pos = -1 start = 0 end = len(index_array) - 1 while start &lt;= end: mid = (start + end) // 2 if find_name == index_array[mid][0]:</description><pubDate>Tue, 21 Jun 2022 06:57:41 GMT</pubDate><category>Major class</category><category>자료구조</category><category>컴퓨터공학과</category><category>검색</category><category>이진검색</category><category>순차검색</category><author>Jisang Lee</author></item><item><title>[ 자료구조 ] 정렬 알고리즘 (Sorting Algorithms)</title><link>https://kkubuck.github.io/notes/tistory-17/</link><guid isPermaLink="true">https://kkubuck.github.io/notes/tistory-17/</guid><description>1. 선택 정렬(앞부터 i, i+1 비교) O(n^2) def selectionSort(array): n = len(array) for i in range(n-1): minIdx = i for k in range(i + 1, n): if (array[minIdx] &gt; array[k]): minIdx = k array[i]</description><pubDate>Tue, 21 Jun 2022 06:56:51 GMT</pubDate><category>Major class</category><category>자료구조</category><category>컴퓨터공학과</category><category>선택정렬</category><category>삽입정렬</category><category>퀵정렬</category><category>버블정렬</category><author>Jisang Lee</author></item><item><title>[ 자료구조 ] 재귀 호출 (recursive call)</title><link>https://kkubuck.github.io/notes/tistory-16/</link><guid isPermaLink="true">https://kkubuck.github.io/notes/tistory-16/</guid><description>1. 재귀호출 이용한 카운트다운 함수 def countDown(n): if n==0: print(&apos;발사!!&apos;) else: print(n) countDown(n-1) 2. 재귀호출 이용한 별 출력 함수 def printStar(n): if n &gt; 0: printStar(n-1) print(&apos;★&apos; * n) 3. 단과 곱할 숫</description><pubDate>Tue, 21 Jun 2022 06:55:12 GMT</pubDate><category>Major class</category><category>자료구조</category><category>컴퓨터공학과</category><category>재귀호출</category><author>Jisang Lee</author></item><item><title>[ 자료구조 ] 그래프 (graph)</title><link>https://kkubuck.github.io/notes/tistory-15/</link><guid isPermaLink="true">https://kkubuck.github.io/notes/tistory-15/</guid><description>1. 그래프 출력 함수 def print_graph(g): for row in range(g.SIZE): for col in range(g.SIZE): print(g.graph[row][col], end = &apos;\\t&apos;) print() print() 2. 깊이 우선 탐색 함수 def find_vertex(g, find_vt</description><pubDate>Tue, 21 Jun 2022 06:54:17 GMT</pubDate><category>Major class</category><category>자료구조</category><category>컴퓨터공학과</category><category>그래프</category><author>Jisang Lee</author></item><item><title>[ 자료구조 ] 이진 트리 (binary tree)</title><link>https://kkubuck.github.io/notes/tistory-14/</link><guid isPermaLink="true">https://kkubuck.github.io/notes/tistory-14/</guid><description>1. 주어진 리스트로 이진 트리 생성 (중복 제거) def generate_book_tree(bookAry): node = TreeNode() node.data = bookAry[0][0] root = node for i in bookAry[1:]: node = TreeNode() node.data= i[0] curren</description><pubDate>Tue, 21 Jun 2022 06:53:19 GMT</pubDate><category>Major class</category><category>자료구조</category><category>컴퓨터공학과</category><category>이진트리</category><author>Jisang Lee</author></item><item><title>[ 자료구조 ] 큐 (queue)</title><link>https://kkubuck.github.io/notes/tistory-13/</link><guid isPermaLink="true">https://kkubuck.github.io/notes/tistory-13/</guid><description>1. 큐가 꽉 차있는지 확인하는 함수 def is_queue_full(self): if self.size - 1 != self.rear: return False elif self.size -1 == self.rear and self.front == -1: return True else: for i in range(self</description><pubDate>Tue, 21 Jun 2022 06:51:29 GMT</pubDate><category>Major class</category><category>자료구조</category><category>컴퓨터공학과</category><category>큐</category><author>Jisang Lee</author></item><item><title>[ 자료구조 ] 스택 (stack)</title><link>https://kkubuck.github.io/notes/tistory-12/</link><guid isPermaLink="true">https://kkubuck.github.io/notes/tistory-12/</guid><description>1. 스택이 비었는지 확인하는 함수 if self.top == self.size - 1: return True else: return False 2. 스택이 꽉 차있는지 확인하는 함수 def is_stack_empty(self): if self.top == -1: return True else: return False 3</description><pubDate>Tue, 21 Jun 2022 06:50:45 GMT</pubDate><category>Major class</category><category>자료구조</category><category>컴퓨터공학과</category><category>스택</category><author>Jisang Lee</author></item><item><title>[ 자료구조 ] 연결리스트 (Linked List)</title><link>https://kkubuck.github.io/notes/tistory-11/</link><guid isPermaLink="true">https://kkubuck.github.io/notes/tistory-11/</guid><description>1. 노드 출력 def printNodes(start): current = start if current == None: return print(current.data, end = &apos; &apos;) while current.link != None: current = current.link print(current.data, end</description><pubDate>Tue, 21 Jun 2022 06:49:32 GMT</pubDate><category>Major class</category><category>자료구조</category><category>컴퓨터공학과</category><category>연결리스트</category><author>Jisang Lee</author></item><item><title>[ 자료구조 ] 선형 리스트 (Linear List)</title><link>https://kkubuck.github.io/notes/tistory-10/</link><guid isPermaLink="true">https://kkubuck.github.io/notes/tistory-10/</guid><description>1. 선형 리스트 생성 languages.append(None) length = len(languages) for i in range(length - 1, position, -1): languages[i] = languages[i - 1] languages[i - 1] = None languages[position] =</description><pubDate>Tue, 21 Jun 2022 06:47:29 GMT</pubDate><category>Major class</category><category>자료구조</category><category>컴퓨터공학과</category><category>선형리스트</category><author>Jisang Lee</author></item></channel></rss>