AI City Challenge CVPR Virtual Workshop 2023

This year’s AI City Challenge workshop will take place on Monday, June 19th at:

East 4, Vancouver Convention Center, 1055 Canada Pl, Vancouver, BC V6C 0C3, Canada

This year’s event will be held in hybrid mode, allowing attendees to participate either in person or remotely. 

For online remote access, paid registrants could find the zoom link from CVPR’s virtual workshop page. Participating teams and invited presenters will receive the zoom join link via email from us organizers. 

Please find the workshop schedule below (Pacific Time):

07:30 AM – 08:00 AM



Breakfast (served from 7:30 to 9:00)

08:00 AM – 08:30 AM



Opening – Workshop Organization Presentation (recording)

08:30 AM – 09:15AM



Keynote (recording)

Topic: Unifying Short and Long-Term Tracking with Graph Hierarchies

Speaker: Laura Leal-Taixe

09:15 AM – 10:10 AM



Paper Presentations – Track1 (7 pre-recorded videos + Q&A)

(1) Enhancing Multi-Camera People Tracking with Anchor-Guided Clustering and Spatio-Temporal Consistency ID Re-Assignment (media)

(2) Multi-camera People Tracking With Mixture of Realistic and Synthetic Knowledge (media)

(3) Integrating Appearance and Spatial-Temporal Information for Multi-Camera People Tracking (media)

(4) ReidTrack: Reid-only Multi-target Multi-camera Tracking (media)

(5) Hierarchical Clustering and Refinement for Generalized Multi-Camera Person Tracking (media)

(6) Addressing the Occlusion Problem in Multi-Camera People Tracking with Human Pose Estimation (media)

(7) Leveraging Future Trajectory Prediction for Multi-Camera People Tracking (media)

10:10 AM – 10:25AM


Morning Coffee (served from 10:00 to 10:45)

10:25 AM – 11:00 AM


Paper Presentations – Track2 (3 pre-recorded videos + Q&A)

(1) Tracked-Vehicle Retrieval by Natural Language Descriptions With Multi-Contextual Adaptive Knowledge (media)

(2) A Unified Multi-modal Structure for Retrieving Tracked Vehicles through Natural Language (media)

(3) Comprehensive Visual Features and Pseudo Labeling for Robust Natural Language-based Vehicle Retrieval (media)

11:00 AM – 12:00 PM



Paper Presentations – Track3 (8 pre-recorded videos + Q&A)

(1) Multi View Action Recognition for Distracted Driver Behavior Localization (media)

(2) Action Probability Calibration for Efficient Naturalistic Driving Action Localization (media)

(3) Multi-Attention Transformer for Naturalistic Driving Action Recognition (media)

(4) Peer-to-Peer Federated Continual Learning for Naturalistic Driving Action Recognition (media)

(5) Multi-View Multi-Scale Driver Action Recognition with Vision Transformer (media)

(6) Triplet Temporal-based Video Recognition with Multiview for Temporal Action Localization (media)

(7) DeepSegmenter: Temporal Action Localization for Detecting Anomalies in Untrimmed Naturalistic Driving Videos (media)

(8) Transformer-based Fusion of 2D-pose and Spatio-temporal Embeddings for Distracted Driver Action Recognition (media)

12:00 PM – 12:50 PM



Lunch (served from 11:45 to 13:30)

12:50 PM – 13:40 PM



Paper Presentations – Track4 (6 pre-recorded videos + Q&A)

(1) Improving Deep Learning-based Automatic Checkout System Using Image Enhancement Techniques (media)

(2) AdaptCD: An Adaptive Target Region-based Commodity Detection System (media)

(3) DACNet: A Deep Automated Checkout Network with Selective Deblurring (media)

(4) Enhancing Retail Checkout through Video Inpainting, YOLOv8 Detection, and DeepSort Tracking

(5) CheckSORT: Refined Synthetic Data Combination and Optimized SORT for Automatic Retail Checkout (media)

(6) Adaptive RoI with pretrained models for Automated Retail Checkout (media)

13:40 PM – 14:30 PM



Paper Presentations – Track5 (6 pre-recorded videos + Q&A)

(1) An Effective Motorcycle Helmet Object Detection Framework for Intelligent Traffic Safety (media)

(2) Robust Automatic Motorcycle Helmet Violation Detection for an Intelligent Transportation System (media)

(3) Helmet Rule Violation Detection for Motorcyclists using a Custom Tracking Framework and Advanced Object Detection Techniques (media)

(4) Video Analytics for Detecting Motorcyclists Helmet Rules Violations (media)

(5) Real-time Multi-Class Helmet Violation Detection Using Few-Shot Data Sampling Technique and YOLOv8 (media)

(6) PRB-FPN+: Video Analytics for Enforcing Motorcycle Helmet Laws (media)

14:30 PM – 15:00 PM



Paper Presentations – Independent (4 pre-recorded videos + Q&A)

(1) Robust and Scalable Vehicle Re-Identification via Self-Supervision (media)

(2) FishEye8K: A Benchmark and Dataset for Fisheye Camera Object Detection (media)

(3) Nordic vehicle dataset (NVD): Performance of vehicle detectors using newly captured NVD from UAV in different snowy weather conditions. (media)

(4) Improving Multi-Agent Motion Prediction with Heuristic Goals and Motion Refinement (media)

15:00 PM – 15:15 PM



Afternoon Coffee Break (served from 15:00 to 15:45)

15:15 PM – 16:15 PM



Panel Discussion on Ethical AI & Synthetic Data Generation (recording)

Panel members:  Shuo Wang, Thomas Tang, David Anastasiu, Ming-Ching, Liang Zheng

16:15 PM – 16:30 PM



Award Ceremony (recording)

16:30 PM – 17:30 PM



Poster Session (@exhibition hall in the west building)

17:30 PM