关注公众号,发现CV技术之美
近日总结了 CVPR 2023 中所有Workshop主要内容(一文了解 CVPR 2023 Workshop 都要做什么),其中不少workshop同时也是有算法竞赛的。
本文汇总所有CVPR 2023中所有的竞赛,总计53个(少数已结束报名),大部分竞赛在进行中或者还未开始,其中不乏特别有商业价值、解决实际问题的比赛,仅其中数据集就非常有价值,不少比赛还提供丰厚奖金,欢迎感兴趣的朋友参赛!
1.Vision-Centric Autonomous Driving (VCAD)
网址:http://vcad.site/#/challenge
技术方向:端到端自动驾驶
赛道
Track 1: OpenLane Topology
Track 2: Online HD Map Construction
Track 3: 3D Occupancy Prediction
Track 4: nuPlan Planning
2.Workshop on Autonomous Driving (WAD)
网址:https://cvpr2023.wad.vision/
赛道
challenge1:Waymo open dataset challenges
2D Video Panoptic Segmentation Challenge、Pose Estimation Challenge、Motion Prediction Challenge、Sim Agents Challenge四项挑战赛,第一名将获得10,000美元的谷歌云奖券。
challenge2:BDD100K challenges
Multiple Object Tracking (MOT)和Multiple Object Tracking and Segmentation (MOTS)两项挑战赛。
challenge3:Argoverse challenges
3.Workshop on End-to-End Autonomous Driving!
网址:https://opendrivelab.com/e2ead/cvpr23
赛道
注:一人可申请多个赛道!
4.The 4th Face Anti-spoofing Workshop and Challenge
网址:https://sites.google.com/view/face-anti-spoofing-challenge/welcome/challengecvpr2023
赛道
5.5th Workshop and Competition on Affective Behavior Analysis in-the-wild
网址:https://ibug.doc.ic.ac.uk/resources/cvpr-2023-5th-abaw/
赛道
6.2nd Workshop on Tracking and Its Many Guises: Tracking Any Object in Open-World
网址:http://taodataset.org/workshop/cvpr23/
截止:5月12日
赛道
7.9th IEEE International Workshop on Computer Vision in Sports (CVsports)
网址:https://www.soccer-net.org/challenges/2023
赛道
Theme 1: Broadcast video Understanding
Theme 2: Field Understanding
Theme 3: Player Understanding
8.4th Embodied AI Workshop
网址:https://embodied-ai.org/
方向:rearrangement, visual navigation, vision-and-language, and audio-visual navigation
9.The 4th CVPR Workshop on 3D Scene Understanding for Vision, Graphics, and Robotics
网址:https://scene-understanding.com/challenge.html
赛道
10.Computer Vision in the Wild
网址:https://computer-vision-in-the-wild.github.io/cvpr-2023/
赛道
11.OmniLabel: Infinite label spaces for semantic understanding via natural language
网址:https://sites.google.com/view/omnilabel-workshop-cvpr23/challenge
截止:4月15日
12.Visual Perception via Learning in an Open World
网址:https://vplow.github.io/vplow_3rd.html#competition
赛道
截止:6月10日
13.2nd Workshop and Challenge on Vision Datasets Understanding
网址:https://sites.google.com/view/vdu-cvpr23/competition
已结束
14.4D Hand Object Interaction: Geometric Understanding and Applications in Dexterous Manipulation
网址:https://4dhoi.github.io/
第一届4D手与物体交互竞赛
截止:5月25日
15.Catch UAVs that Want to Watch You: Detection and Tracking of Unmanned Aerial Vehicle (UAV) in the Wild and the 3rd Anti-UAV Workshop & Challenge
网址:https://anti-uav.github.io/Evaluate/
赛道
已结束
16.RetailVision - Revolutionizing the World of Retail
网址:https://retailvisionworkshop.github.io/#challenge
方向:商品检索
截止:5月31日
赛道
17.VAND: Visual Anomaly and Novelty Detection
网址:https://sites.google.com/view/vand-cvpr23/home
方向:工业视觉检测
时间:第一阶段:3月13日至4月14日;第二阶段:4月17日-5月12日
赛道
18.Woodscape Challenge
网址:https://sites.google.com/view/omnicv2023/challenges/woodscape-challenge
方向:实时处理鱼眼相机图像,行人和车辆检测、车道标记检测、路边检测等。
19.Multi-View Layout Challenge
网址:https://sites.google.com/view/omnicv2023/challenges/multi-view-layout-challenge
方向:使用配准的360图像序列作为唯一输入的房间布局估计解决方案
截止:6月3日
20.Odometry Challenge using Omnidirectional Drone Data
网址:https://sites.google.com/view/omnicv2023/challenges/omnidirectional-odometry-challenge
方向:开发一种里程计算法,利用一年中不同季节拍摄的鱼眼相机图像和激光雷达点云数据集,准确跟踪无人机在森林环境中的运动。该算法可以是基于视觉的,也可以是基于激光雷达的,还可以是两种传感器的融合。
截止:TBD
奖金:获奖团队/个人可获得1000美元
21.2023 Challenge onzero-shot image captioning evaluation
网址:https://nice.lgresearch.ai/
方向:零样本图像字幕
奖金:每赛道的冠军将获得5000美元的奖金,亚军将获得2500美元
截止:4月30日
赛道
22.7th AI City Challenge Workshop
网址:https://www.aicitychallenge.org/2023-challenge-tracks/
23.Joint 3rd Ego4D and 11th EPIC Workshop on Egocentric Vision
网址:https://sites.google.com/view/ego4d-epic-cvpr2023-workshop/
Ego4D Challenge 2023 举办14项挑战,包括2项新的挑战(EgoTracks和PACO Zero-Shot)。
网址:https://ego4d-data.org/docs/challenge/#episodic-memory
奖金:第一名1500美元;第二名1000美元;第三名500美元
截止:5月19日
赛道
Visual queries with 2D localization (VQ2D) and Visual Queries 3D localization (VQ3D)
Natural language queries (NLQ)
Moments queries (MQ)
EgoTracks
PACO Zero-Shot
Temporal localization
Object state change classification
Audio-visual speaker diarization
Speech transcription
Talking to me
Looking at me
Short-term hand object prediction
Long-term activity prediction
25.Pixel-level Video Understanding in the Wild Challenge
网址:https://www.vspwdataset.com/Workshop%202023.html
赛道
26.First Rhobin Challenge - Reconstruction of human-object interactionXi Wang
网址:https://rhobin-challenge.github.io/
方向:三维人体重建、6DoF姿态估计、人和物体的联合重建
赛道
网址:https://codalab.lisn.upsaclay.fr/competitions/9280
截止:5月5日
27.ACDC CHALLENGE 2022
网址:https://acdc.vision.ee.ethz.ch/news
在恶劣的视觉条件下对驾驶场景进行语义理解
方向:
Normal-to-adverse domain adaptation on Cityscapes to ACDC
Supervised semantic segmentation in adverse conditions
Supervised semantic segmentation in individual adverse conditions
Uncertainty-aware semantic segmentation
28.Fourth workshop Neural Architecture Search Third lightweight NAS challenge
网址:https://cvpr-nas.com/competition
方向:神经架构搜索(NAS)
29.3rd Mobile AI Workshop and Challenges
网址:https://ai-benchmark.com/workshops/mai/2023/
挑战赛:TBD
30.Workshop on Foundation Models1st foundation model challenge
To be Updated
31.Meta AI Video Similarity Challenge
网址:https://www.drivendata.org/competitions/group/meta-video-similarity/
赛道
32.The 3rd Workshop of Adversarial Machine Learningon Computer Vision: Art of Robustness
网址:https://robustart.github.io/#challenge
挑战赛:TBD
33.L3D-IVU: 2nd Workshop on Learning with Limited Labelled Data for Image and Video Understanding
网址:https://sites.google.com/view/l3d-ivu-2023
赛道
34.SnakeCLEF2023
网址:https://sites.google.com/view/fgvc10/competitions/snakeclef2023
方向:蛇物种分类
截止:5月17日
35.GeoLifeCLEF2023
网址:https://sites.google.com/view/fgvc10/competitions/geolifeclef2023
方向:利用各种可能的预测因素来预测特定地点和时间内的植物
截止:5月17日
36.FathomNet2023
网址:https://sites.google.com/view/fgvc10/competitions/fathomnet2023
方向:海洋生物物种识别
截止:5月23日
37.PlantTraits2023
网址:https://sites.google.com/view/fgvc10/competitions/planttraits
方向:使用基于深度学习的回归模型(如CNN)从植物照片中预测植物性状。
截止:5月24日
38.iBioHash2023
网址:https://sites.google.com/view/fgvc10/competitions/ibiohash2023
方向:基于learning-to-hash技术解决大规模的细粒度图像搜索任务
截止:5月15日
39.4th Workshop on Continual Learning in Computer Vision (CLVision)
网址:https://sites.google.com/view/clvision2023/challenge
方向:持续学习
截止:5月20日
40.BuildingNet Challenge (CVPRW)
网址:https://eval.ai/web/challenges/challenge-page/1938/overview
截止:5月24日
41.The SkelNetOn Challenge (2D)
网址:https://sites.google.com/view/dlgc-workshop-cvpr2023/challenge
SkelNetOn
挑战赛是围绕四个领域的形状理解而展开的。平台提供形状数据集和一些补充资源(如前/后处理、采样和数据增强的脚本)以及测试平台。
已结束
42.3DCoMPaT Challenge
网址:https://3dcompat-dataset.org/workshop/#main-section
方向:三维分割
43.Workshop on Vision-based InduStrial InspectiON (VISION)
网址:https://vision-based-industrial-inspection.github.io/cvpr-2023/
方向:基于视觉的工业检测
截止:5月1日
赛道
44.3D Surface Mesh Estimation Challenge
网址:https://www.cvpr2023-dl-ultrasound.com/single-project
方向:工业超声图像分析
奖品:第一名5000CAD;第二名3000CAD;第三名1000CAD
截止:5月30日
45.3rd International Scan-to-BIM challenge
网址:https://cv4aec.github.io/
截止:TBD
赛道
46.The Amazon Rainforest Challenge
网址:https://sites.google.com/view/rainforest-challenge/multiearth-2023
Tasks to be released
47.VizWiz Grand Challenge: Describing Images and Videos Taken by Blind People
该组会发起四个挑战赛,如下:
visual question answering challenge task
网址:https://vizwiz.org/tasks-and-datasets/vqa/
answer grounding challenge
网址:https://vizwiz.org/tasks-and-datasets/answer-grounding-for-vqa/
salient object detection challenge
网址:https://vizwiz.org/tasks-and-datasets/salient-object-detection/
few-shot object recognition challenge
网址:https://eval.ai/web/challenges/challenge-page/1896/overview
截止:5月12日
48.AVA: Accessibility, Vision, and Autonomy Meet
网址:https://accessibility-cv.github.io/index.html#challenge
方向:无障碍人工智能的研究,如无障碍住宅、医院、城市、基础设施、交通。
奖项:组会演讲、500美元的资金奖励和一台OAK-D相机
截止:11月6日
49.19th CVPR Workshop on Perception Beyond the Visible Spectrum (PBVS 2023)
网址:https://pbvs-workshop.github.io/challenge.html
已结束
50.The 3rd Workshop on Light Fields for Computer Vision LFNAT: New Applications and Trends in Light Fields
网址:http://www.lfchallenge.com/LFNAT@CVPR2023/challenge/
方向:捕捉语义分割和光场深度估计方面的先进技术
赛道
51.Second Workshop of Mobile Intelligent Photography and Imaging
网址:http://mipi-challenge.org/MIPI2023/
已结束
52.8th New Trends in Image Restoration and Enhancement Workshop and Challenges
网址:https://cvlai.net/ntire/2023/
方向:
Image/video inpainting
Image/video deblurring
Image/video denoising
Image/video upsampling and super-resolution
Image/video filtering
Demosaicing
Image/video compression
Image/video de-hazing, de-raining, de-snowing, etc.
53.AVAILABLE CHALLENGES
网址:http://cvpr2023.ug2challenge.org/#
赛道
网址:http://cvpr2023.ug2challenge.org/track1.html
目标:评估和推进目标检测算法对有雾图像的鲁棒性
网址:http://cvpr2023.ug2challenge.org/track2.html
目标:开发新的图像重建算法,用于通过非平面湍流的非相干成像。
子赛道:
Text Recognition through Atmospheric Turbulence
Coded Target Restoration through Atmospheric Turbulence
网址:http://cvpr2023.ug2challenge.org/track3.html
方向:图像去雨
END
欢迎加入「竞赛」交流群备注:竞赛