task_categorization.md

Current task types

This list shows how we reformulating existing visual perception tasks. We are working on it to support more task types.

Detection

• Visual Grounding
• Object Detection
• 2D Object Detection
• Small Object Detection
• Defect Detection
• Face Detection
• License Plate Detection
• Anomaly Detection
• Human Detection
• Surgical Tool Detection
• Dense Object Detection
• Open World Object Detection
• Zero-Shot Object Detection
• Animal Action Recognition
• Robotic Grasping
• Object Localization
• Hand Detection
• Visual Relationship Detection
• Open Vocabulary Object Detection
• Oriented Object Detection
• Object Detection in Indoor Scenes
• Object Detection in Aerial Images
• Person Search
• Object Recognition

Segmentation

• Semantic Segmentation
• Instance Segmentation
• Lane Detection
• 2D Semantic Segmentation
• Medical Image Segmentation
• Human Part Segmentation
• Action Segmentation
• Video Object Segmentation
• Referring Expression Segmentation
• Saliency Detection
• Salient Object Detection
• The Semantic Segmentation of Remote Sensing Imagery
• Crack Segmentation
• Action Unit Detection
• RGB Salient Object Detection
• Boundary Detection
• Crack Segmentation for Infrastructure
• Surgical Tool Segmentation

Counting

• Object Counting
• Crowd Counting
• Density Estimation
• Pedestrian Detection
• Crowd Estimation in Dense Scenes
• Traffic Counting in Surveillance

VQA

• Visual Question Answering (VQA)
• Classification
• Image Captioning
• Question Answering
• Visual Reasoning
• Visual Question Answering
• Relational Reasoning