ConvNeXt- A ConvNet for the 2020s - Paper Explained
Offered By: Aleksa Gordić - The AI Epiphany via YouTube
Course Description
Overview
Explore a comprehensive analysis of the "A ConvNet for the 2020s" paper in this 40-minute video lecture. Delve into the convergence of transformers and CNNs, understand the main diagram and its corrections, and recap the Swin transformer. Learn about modernizing ResNets, dive deeper into stage ratios and miscellaneous topics like inverted bottlenecks and depthwise convolutions. Examine the results in classification, object detection, and segmentation tasks. Gain insights into how ConvNets outperform vision transformers in big data regimes without attention layers, demonstrating the enduring relevance of convolutional priors in computer vision.
Syllabus
Intro - convergence of transformers and CNNs
Main diagram explained
Main diagram corrections
Swin transformer recap
Modernizing ResNets
Diving deeper: stage ratio
Diving deeper: misc inverted bottleneck, depthwise conv...
Results classification, object detection, segmentation
RIP DanNet
Summary and outro
Taught by
Aleksa Gordić - The AI Epiphany
Related Courses
Vision Transformers Explained + Fine-Tuning in PythonJames Briggs via YouTube Do Vision Transformers See Like Convolutional Neural Networks - Paper Explained
Aleksa Gordić - The AI Epiphany via YouTube Stable Diffusion and Friends - High-Resolution Image Synthesis via Two-Stage Generative Models
HuggingFace via YouTube Intro to Dense Vectors for NLP and Vision
James Briggs via YouTube Geo-localization Framework for Real-world Scenarios - Defense Presentation
University of Central Florida via YouTube