Looking 3D: Anomaly Detection with 2D-3D Alignment
A novel task of Conditional Anomaly Detection (AD), where the goal is to detect anomalies in a query based on a reference 3D shape.
Research
Improving Semantic Correspondence with Viewpoint-Guided Spherical Maps
A technique to improve semantic corresponding matching through 3D prior
Who are you referring to? Coreference resolution in image narrations
A novel task of resolving coreferences in image narrations. For benchmarking and enabling the progress, we introduce a dataset – CIN – that contains images w...
Learning Multiple Dense Prediction Tasks from Partially Annotated Data
A more realistic and general setting for MTL and an architecture-agnostic algorithm to learn MTL on partially annotated data.
Not All Relations are Equal: Mining Informative Labels for Scene Graph Generation
A novel model-agnostic training framework for scene graph generation based on the concept of label informativeness.
Cross-domain Few-shot Learning with Task-specific Adapters
A task-adaptation technique and systematic analysis for cross-domain few-shot classification.
Universal Representation Learning from Multiple Domains for Few-shot Classification
A universal representation learning technique for cross-domain few-shot classification.
Dataset Condensation with Gradient Matching
A training set synthesis technique for condensing a large dataset into a small set of informative synthetic samples.