Schedule
Program Schedule
8:30 AM - 8:40 AM Opening remarks
8:40 AM - 9:00 AM Luca Carlone Â
9:00 AM - 9:05 AM Q&A
9:05 AM - 9:25 AM Nicolas Padoy
9:25 AM - 9:30 AM Q&A
9:30 AM - 9:50 AM Helisa Dhamo
9:50 AM - 9:55 AM Q&A
10:00 AM - 10:30 AM Spotlights
10:30 AM - 10:55 AM Coffee break + Poster session
10:55 AM - 11:55 AM Oral sessionÂ
11:55 AM - 12:00 PM Break
12:00 PM - 12:20 PM Ranjay Krishna
12:20 PM - 12:25 PM Q&A
12:25 PM - 12:45 PM Bernard Ghanem
12:45 PM - 12:50 PM Q&A
12:50 PM - 1:00 PM Concluding remarks and Best Paper Award
Presentations Schedule
10:00 AM - 10:03 AM DeepCut: Unsupervised Segmentation using Graph Neural Networks Clustering
10:03 AM - 10:06 AM Relational Prior Knowledge Graphs for Detection and Instance Segmentation
10:06 AM - 10:09 AM Polygon Detection for Room Layout Estimation using Heterogeneous Graphs and Wireframes
10:09 AM - 10:12 AM Detecting Objects with Context-Likelihood Graphs and Graph Refinement
10:12 AM - 10:15 AM Incremental Object Grounding Using Scene Graphs
10:15 AM - 10:18 AM Auto3DSG: Autoencoding for 3D Scene Graph Learning via Object-Level Scene Reconstruction
10:18 AM - 10:21 AM Graph-based 3DoF Camera Localization on Object Maps
10:21 AM - 10:24 AM Haystack: A Panoptic Scene Graph Dataset to Evaluate Rare Predicate Classes
10:24 AM - 10:27 AM Exploring the Road Graph in Trajectory Forecasting for Autonomous Driving
10:27 AM - 10:30 AM nuScenes Knowledge Graph - A comprehensive semantic representation of traffic scenes for trajectory prediction
10:30 AM - 10:55 AM Coffee break + Poster session (all accepted works are welcome to bring a poster)
10:55 AM - 11:05 AM Dynamic Scene Graph Representation for Surgical Video
11:05 AM - 11:10 AM Q&A
11:10 AM - 11:20 AM Knowledge Informed Sequential Scene Graph Verification Using VQA
11:20 AM - 11:25 AM Q&A
11:25 AM - 11:35 AM Fine-Grained is Too Coarse: A Novel Data-Centric Approach for Efficient Scene Graph Generation
11:35 AM - 11:40 AM Q&A
11:40 AM - 11:50 AM SceneGenie: Scene Graph Guided Diffusion Models for Image Synthesis
11:50 AM - 11:55 AM Q&A