Computer Vision with PyTorch
summary
Covers concrete workflows using PyTorch and torchvision to run pretrained models like Mask R-CNN on COCO images, producing bounding boxes with confidence scores (examples show ~99%, ~80%, down into the 30s); notebooks live on GitHub and are runnable in Google Colab with GPU/CUDA; demonstrates training loops, datasets (CIFAR‑10, COCO, ImageNet), fine‑tuning ResNet backbones, and cites ImageNet top‑5 error around ImageNet 1.8%, so you see code, expected metrics, and where GPUs matter.
AI-written summary · about
chapters (10)
- 0:00Computer vision intro
- 2:00What is computer vision
- 8:40Why CNNs excel
- 10:30PyTorch basics & GPU
- 17:58Working with images
- 23:28Pre-trained models
- 32:00Training models from scratch
- 37:02Building image classifier
- 49:54Summary & resources
- 51:19Q&A
related talks
01h 27mHands-On Computer Vision with PyTorch
01h 29mComputer Vision: An (Un?)Expected Journey, with Keras and Tensorflow
01h 39mGoogle Colaboratory and Python: Deep Learning basics for CV tasks
01h 32mIntroduction to PyTorch
36mObject detection to Instance segmentation: Learn to apply several algorithms...
22mObject Detection with KerasCV
47mDeep Learning with PyTorch
01h 15mTrain Object Detection with Small Datasets