Detection
Also written: Object Detection
Identifying and locating specific objects within images or videos.
Detection is about finding and pinpointing specific things in pictures or videos. Think of it like a computer looking at a photo and drawing a box around every car it sees.
This is a core task in computer vision, used for everything from self-driving cars identifying pedestrians to social media apps recognizing faces.
Detection in computer vision involves identifying the presence and location of instances of predefined object classes within an image or video. This is typically achieved by outputting bounding boxes around detected objects along with a class label and a confidence score.
Modern detection models often leverage deep convolutional neural networks, with architectures like Faster R-CNN, YOLO, and SSD being prominent. Tradeoffs exist between accuracy, speed, and the ability to detect small or overlapping objects.
Meta AI's recent work on SAM 3 highlights the ongoing effort to unify different computer vision tasks, including detection and tracking, into single, more capable models. This integration aims to improve efficiency and performance across a range of visual understanding applications.
- SAM 3: Building a unified model architecture for detection and trackingMeta · AI at Meta on YouTube ·
- Meta Meta AI
- Meta AI is developing unified models like SAM 3 that integrate detection capabilities with other visual tasks.
- Segmentation
- Segmentation identifies the precise pixel boundaries of an object,
- Classification
- Classification assigns a label to an entire image, without locat
Object detection has roots in early computer vision research, with significant advancements driven by deep learning techniques starting in the early 2010s. Source ↗
Use an online image analysis tool that offers object detection. Upload a photo and see if it can identify common objects like people, cars, or animals, drawing boxes around them.
The Playbook: reusable how-tos, prompts, configs and tips pulled out of the news.
- PromptRole-play a student struggling with math for AI tutor simulationSimulate a student's math struggles and uncertainty for AI tutor evaluation. · 5 min
- TemplateReusable workflow templates for healthcare teamsTeams can create and share templates for common workflows like documentation or prior authorizations. · 15 min
- TemplateTemplate for building an AI tool tailored to your workA structured approach to defining and building a custom AI tool for specific business needs. · 15 min
- TemplateEvals CSV prompt list for agent skillsA CSV file template for listing prompts used in evaluating agent skills. · 20 min
- HeuristicDouble coding agent task completion with docs linksProviding coding agents with a documentation link significantly increases their task completion rate. · 2 min
Questions people ask
- What is the difference between detection and classification?
- Classification tells you what is in an image overall, while detection tells you what objects are present and where they are located using bounding boxes.
- What is a bounding box in detection?
- A bounding box is a rectangular box drawn around a detected object in an image to indicate its position and extent.
First seen on Subvolts Dec 8, 2025. Updated Sep 15, 2026. Written by an AI model from the items above and reviewed against them weekly; report an error.