Explore zero shot object detection and localization using OpenAI CLIP in this 30-minute video tutorial. Learn about the evolution of computer vision, from the ILSVRC competition to the recent advancements in multi-modal models. Discover how to implement zero shot object detection without fine-tuning, saving time and resources. Follow along as the tutorial covers data preprocessing, initializing CLIP in Python, applying localization techniques, and creating bounding boxes with Matplotlib. Gain insights into the latest trends in multi-modal machine learning and witness the power of CLIP in performing object detection across different domains without additional training.
Overview
Syllabus
Early Progress in Computer Vision
Classification vs. Localization and Detection
Zero Shot with OpenAI CLIP
Zero Shot Object Localization with OpenAI CLIP
Localization with Occlusion Algorithm
Zero Shot Object Detection with OpenAI CLIP
Data Preprocessing for CLIP
Initializing OpenAI CLIP in Python
Clipping the Localization Visual
Applying Scores for Visual
Object Localization with New Prompt
Zero Shot Object Detection in Python
Creating Bounding Boxes with Matplotlib
Object Detection Code
Object Detection Results
Trends in Multi-Modal ML
Taught by
James Briggs
Related Courses
-
Object Localization with TensorFlow
-
Computer Vision: YOLO Custom Object Detection with Colab GPU
-
OpenAI CLIP for Zero-Shot Image Classification and Image Similarity - Lab Tutorial
-
OpenAI CLIP Explained - Multi-modal ML
-
Zero-Shot Image Classification with OpenAI's CLIP Model
-
Object and Bounding Box Detection with Vision Models Using Florence 2
