CV计算机视觉每日开源代码Paper with code速览-2023.10.10

精华置顶

墙裂推荐！小白如何1个月系统学习CV核心知识：链接

点击@CV计算机视觉，关注更多CV干货

论文已打包，点击进入—>下载界面

点击加入—>CV计算机视觉交流群

1.【基础网络架构：Transformer】Hierarchical Side-Tuning for Vision Transformers

• 论文地址：https://arxiv.org//pdf/2310.05393

• 开源代码（即将开源）：https://github.com/AFeng-x/HST

2.【目标检测】Anchor-Intermediate Detector: Decoupling and Coupling Bounding Boxes for Accurate Object Detection

• 论文地址：https://arxiv.org//pdf/2310.05666

• 开源代码（即将开源）：https://github.com/YilongLv/AID

3.【目标检测】How to effectively train an ensemble of Faster R-CNN object detectors to quantify uncertainty

• 论文地址：https://arxiv.org//pdf/2310.04829

• 开源代码：https://github.com/Akola-Mbey-Denis/EfficientEnsemble

4.【目标检测】HalluciDet: Hallucinating RGB Modality for Person Detection Through Privileged Information

• 论文地址：https://arxiv.org//pdf/2310.04662

• 代码即将开源

5.【语义分割】Towards Dynamic and Small Objects Refinement for Unsupervised Domain Adaptative Nighttime Semantic Segmentation

• 论文地址：https://arxiv.org//pdf/2310.04747

• 工程主页：DSRNSS

• 代码即将开源

6.【目标跟踪】Lightweight Full-Convolutional Siamese Tracker

• 论文地址：https://arxiv.org//pdf/2310.05392

• 开源代码（即将开源）：https://github.com/LiYunfengLYF/LightFC

7.【3D目标检测】Towards Fair and Comprehensive Comparisons for Image-Based 3D Object Detection

• 论文地址：https://arxiv.org//pdf/2310.05447

• 开源代码（即将开源）：https://github.com/OpenGVLab/3dodi

8.【点云】Bidirectional Knowledge Reconfiguration for Lightweight Point Cloud Analysis

• 论文地址：https://arxiv.org//pdf/2310.05125

• 开源代码（即将开源）：https://github.com/cuixing100876/BKR

9.【点云3D目标检测】Uni3DETR: Unified 3D Detection Transformer

• 论文地址：https://arxiv.org//pdf/2310.05699

• 开源代码：https://github.com/zhenyuw16/Uni3DETR

10.【点云3D目标检测】Fully Sparse Long Range 3D Object Detection Using Range Experts and Multimodal Virtual Points

• 论文地址：https://arxiv.org//pdf/2310.04800

• 开源代码（即将开源）：https://github.com/ajinkyakhoche/RangeFSD

11.【医学图像分割】Cross-head mutual Mean-Teaching for semi-supervised medical image segmentation

• 论文地址：https://arxiv.org//pdf/2310.05082

• 开源代码（即将开源）：https://github.com/Leesoon1984/CMMT-Net

12.【医学图像分割】A Simple and Robust Framework for Cross-Modality Medical Image Segmentation applied to Vision Transformers

• 论文地址：https://arxiv.org//pdf/2310.05572

• 开源代码：https://github.com/matteo-bastico/MI-Seg

13.【医学图像分割：3D】HartleyMHA: Self-Attention in Frequency Domain for Resolution-Robust and Parameter-Efficient 3D Image Segmentation

• 论文地址：https://arxiv.org//pdf/2310.04466

• 开源代码：https://github.com/IBM/multimodal-3d-image-segmentation

14.【多模态】FLATTEN: optical FLow-guided ATTENtion for consistent text-to-video editing

• 论文地址：https://arxiv.org//pdf/2310.05922

• 工程主页：FLATTEN Video Editing

• 开源代码（即将开源）：https://github.com/yrcong/FLATTEN_video_editing/

15.【多模态】Interpreting CLIP's Image Representation via Text-Based Decomposition

• 论文地址：https://arxiv.org//pdf/2310.05916

• 工程主页：Interpreting CLIP's Image Representation via Text-Based Decomposition

• 开源代码：https://github.com/yossigandelsman/clip_prs

16.【多模态】UReader: Universal OCR-free Visually-situated Language Understanding with Multimodal Large Language Model

• 论文地址：https://arxiv.org//pdf/2310.05126

• 开源代码（即将开源）：https://github.com/LukeForeverYoung/UReader

17.【多模态】Building an Open-Vocabulary Video CLIP Model with Better Architectures, Optimization and Data

• 论文地址：https://arxiv.org//pdf/2310.05010

• 开源代码：https://github.com/wengzejia1/Open-VCLIP

18.【多模态】Symmetrical Linguistic Feature Distillation with CLIP for Scene Text Recognition

• 论文地址：https://arxiv.org//pdf/2310.04999

• 开源代码：https://github.com/wzx99/CLIPOCR

19.【多模态】HowToCaption: Prompting LLMs to Transform Video Annotations at Scale

• 论文地址：https://arxiv.org//pdf/2310.04900

• 开源代码：https://github.com/ninatu/howtocaption

20.【多模态】Rephrase, Augment, Reason: Visual Grounding of Questions for Vision-Language Models

• 论文地址：https://arxiv.org//pdf/2310.05861

• 开源代码：https://github.com/archiki/RepARe

21.【自动驾驶：深度估计】Federated Self-Supervised Learning of Monocular Depth Estimators for Autonomous Vehicles

• 论文地址：https://arxiv.org//pdf/2310.04837

• 开源代码：https://github.com/eltonfss/federated-sc-depth

22.【轨迹预测】SocialCircle: Learning the Angle-based Social Interaction Representation for Pedestrian Trajectory Prediction

• 论文地址：https://arxiv.org//pdf/2310.05370

• 开源代码：https://github.com/cocoon2wong/SocialCircle

23.【图像检索】Sentence-level Prompts Benefit Composed Image Retrieval

• 论文地址：https://arxiv.org//pdf/2310.05473

• 开源代码（即将开源）：https://github.com/chunmeifeng/SPRC

论文已打包，点击进入—>下载界面

CV计算机视觉交流群

群内包含目标检测、图像分割、目标跟踪、Transformer、多模态、NeRF、GAN、缺陷检测、显著目标检测、关键点检测、超分辨率重建、SLAM、人脸、OCR、生物医学图像、三维重建、姿态估计、自动驾驶感知、深度估计、视频理解、行为识别、图像去雾、图像去雨、图像修复、图像检索、车道线检测、点云目标检测、点云分割、图像压缩、运动预测、神经网络量化、网络部署等多个领域的大佬，不定期分享技术知识、面试技巧和内推招聘信息。

想进群的同学请添加微信号联系管理员：PingShanHai666。添加好友时请备注：学校/公司+研究方向+昵称。

推荐阅读：

CV计算机视觉每日开源代码Paper with code速览-2023.10.9

CV计算机视觉每日开源代码Paper with code速览-2023.10.6