Overview

Aims and Scope

The VISION workshop aims to provide a platform for the exchange of scholarly innovations and emerging practical challenges in Vision-based Industrial Inspection. Through a series of keynote talks, technical presentations, and data challenges, this workshop is intended to (i) bring together researchers from the interdisciplinary research communities related to computer vision-based inspection; (ii) connect researchers and industry practitioners to synergize recent research progress and current needs in industrial practice.

Guest Speakers

Announcement

Best Paper

Segment Anything Is Not Always Perfect: An Investigation of SAM on Different Real-world Applications

Wei Ji et al. (University of Alberta, University of Amsterdam, Dalian University of Technology, Samsung Research America)

Challenge Competition

Track	1st place	2nd place	3rd place	Most Innovative awards
Track 1	Lusterinc Inc	Chinese Academy of Science	Micro-i Inc	IE Vision Inc & SYS University
Track 2	LeadTech & Sichuan University	Dinnar Inc	FS-Tech Inc	Chinese Academy of Science

Program

Important Notice to Presenters

• Please confirm your attendance type: in-person or virtual.
• Please submit your slides and posters before June 14th, EOD. (poster presenters only need to submit their poster, oral presenters need to submit both).
• If you plan to give a virtual presentation, please submit your recorded video before June 14th, EOD. Please note that registration is still needed for virtual attendance.
• Please note that the slides, poster, and recording will be published online and must be the final version.
• All the accepted oral presentation papers are also invited for a poster presentation.
• Poster specs can be found here.
• A Zoom link will be provided to the virtual presentation teams upon request.

Time(PDT)	Topic
8:30 - 8:45	Opening Remarks Rahul Kapoor
8:45 - 9:15	Keynote: Machine Vision Enabled In-Process Quality Improvement in Smart Manufacturing Dr. Jianjun Shi Download PDF
9:15 - 9:45	Keynote: Video-image data mining for zero-waste additive manufacturing Dr. Bianca Maria Colosimo Download PDF
9:45 - 10:00	Coffee Break
10:00 - 12:00	Best Paper Announcement and Paper Presentation Towards Automated Polyp Segmentation Using Weakly- and Semi-Supervised Learning and Deformable Transformers Guangyu Ren et al. (Imperial College London) Download PDF Download PPT XDNet: A Few-Shot Meta-Learning Approach for Cross-Domain Visual Inspection Lasitha S Vidyaratne et al. (Industrial AI Lab, Hitachi America, Ltd. R&D) Download PPT Glass Wool Defect Detection Using an Improved YOLOv5 Yizhou Jin et al. (Beihang University, Huazhong University of Science & Technology) Download PDF How Do Label Errors Affect Thin Crack Detection by DNNs Liang Xu et al. (Tohoku University) Download PDF Synthetic Data for Defect Segmentation on Complex Metal Surfaces Juraj Fulir et al. (Fraunhofer ITWM, RPTU Kaiserslautern-Landau) Download PDF Download PPT Diversified and Multi-Class Controllable Industrial Defect Synthesis for Data Augmentation and Transfer Jing Wei et al. (Chinese Academy of Sciences; University of Chinese Academy of Sciences) Download PPT Segment Anything Is Not Always Perfect: An Investigation of SAM on Different Real-world Applications Wei Ji et al. (University of Alberta, University of Amsterdam, Dalian University of Technology, Samsung Research America) Download PDF Download PPT Hierarchical Dense Correlation Distillation for Few-Shot Segmentation Bohao Peng; Zhuotao Tian (The Chinese University of Hong Kong)*; Shu Liu (SmartMore) Download PDF Set Features for Fine-grained Anomaly Detection Niv Cohen et al. (The Hebrew University of Jerusalem) Download PDF Download PPT
12:00 - 14:00	Lunch Break and Poster Session@exhibit hall in the west building A dataset to recognize any material anywhere_Manuel Drehwald.zip Assigned MURA Defect Generation Based on Diffusion Model - LiuQiang.zip Attention Modules Improve Image Level Anomaly Detection A Differ Net Case Study - Andre Vieirae Silva.zip Back to the Feature Classical 3D Features are (Almost) All You Need for 3D Anomaly Detection- Eliahu Horwitz.pdf.zip Data-Centric AI is the Key to Democratization and Scalability - Yong Park.zip Leveraging Multi-view Data for Improved Detection Performance_ An Industrial Use Case_Vivasvan Patel - Faranak Shamsafar.zip Leveraging Noisy Boundaries for Semi Supervised Industrial Defect Detection - Santiago Wang.zip N-pad Neighboring Pixel-based Industrial Anomaly Detection - 장준규.pdf.zip Parcel3D Shape Reconstruction from Single RGB Images for Applications in Transportation Logistics - Alexander Naumann.pdf.zip Revisiting Deep Video Motion Magnification for Real-time Applications - Hyunwoo Ha.zip Towards Sim-to-Real Industrial Parts Classification with Synthetic Dataset - XiaomengZhu.pdf.zip Unsupervised Automatic Defect Inspection based on Image Matching and Local One-class Classification - Chengkan Lv.zip What Makes A Good Data Augmentation For Few-Shot Unsupervised Image Anomaly Detection - Shuheng Zhang.zip
14:00 - 14:30	Keynote: Efficient and Less-biased Visual Learning Dr. Leonid Sigal Download PDF
14:30 - 15:00	Keynote: Imaging under challenging conditions – an application case in steel mills Dr. Tzyy-Shuh Chang Download PDF
15:00 - 15:45	Panel Discussion Moderator: Dr. Cinbis (METU) Panelists: Dr. Jianjun Shi (Georgia Tech), Dr. Bianca Colosimo (Politecnico di Milano), Dr. Leonid Sigal (UBC), Dr. Tzyy-Shuh Chang (OG Technology), Rahul Kapoor (Apple), Joseph Robin (Apple), and Dr. Danny Bickson (Visual Layer)
15:45 - 16:00	Coffee Break
16:00 - 18:00	Data Competition Presentation Message from Data Challenge Organizers Data Challenge organziers Lusterinc team solution presentation LUSTER LightTech Co. Download PPT Aoi-overfitting team solution presentation CASI Vision Technology Co., Ltd.; Institute of Automation, Chinese Academy of Sciences. Download PPT Micro-i team solution presentation Microintelligence Research. Download PDF IEVision&SYSU team solution presentation Intelligent Eyes (IEVision) Inc.; Sun Yat-sen University. Download PPT Q&A Leadtech-SCU Data Gen team solution presentation Leadtech Co. Ltd; Sichuan University. Download PPT DINNAR_RD team solution presentation Dinnar; XJTLU. Download PPT FS tech team solution presentation FS tech. Download PDF CAS-ELIIV-T2 team solution presentation Institute of Automation, Chinese Academy of Sciences; CASI Vision Technology CO. Download PPT Q&A

Call for Papers

Submission Instructions

Authors have two options:

• Regular paper: max 8 pages (including figures and tables), appear in the CVPR workshops proceedings
• Extended Abstracts: 4 pages, will NOT appear in the proceedings

Topics of Interest

From an industry point of view:

• Opportunities for Development in vision-based inspection technology, new problem formulation, data formats, and potential applications
• Deployment Practices and Challenges of production vision-based inspection systems
• Open Source Datasets for vision-based inspection research

From an academic point of view:

• Unsupervised, Semi-supervised, and Supervised Anomaly Detection under data and annotation limitations
• Data Generation Techniques for small datasets and rare defect types
• Data-centric Tools and Methodologies for more suitable data collection procedures, higher data quality, and efficiency
• Robust, Explainable Machine Learning for high-stake applications
• Vision-based modeling, diagnosis, control, and quality improvement of the manufacturing system

Important Dates

March 19th (Extended)	Paper Submission Deadline
April 5th	Author Notification
April 14th	Camera-ready

May 7th (Extended)	Paper Submission Deadline
April 15th – May 14th	Rolling Decisions

Challenges

Participation

The detailed instructions for the data challenge and the dataset can be found via the submission links.

Track	Description	Make a Challenge Submission
Challenge 1	Data-efficient Defect Detection
Challenge 2	Data-generation for Defect Detection

Important Dates

Jan 30th	Dataset Release and Registration Open
May 1st	Result Submission Deadline
May 7th	Technical Report Submission Deadline
May 20th	Winners Announcement

Organizers

Workshop Organizers

Challenge Organizers

Advisory Committee

Follow us