Visualizing image embeddings in a 2D space
- Installing Python 3.11
- Installing FiftyOne with venv
- Running inference notebooks inside a Dev Container
- Setting up GPU and TPU usage on Kaggle Noteboks
- Setting up a wandb API token and using it on Kaggle Notebooks and Google Colab
- Setting up kagglehub on Google Colab
- Setting up a HuggingFace token
- Copying data folders without download through Google Drive
- Digital Image Representation in PIL, NumPy, and PyTorch
- Training a Multilayer Perceptron for Image-based Regression
- Matrix Multiplication, Non-linear Activations, and Network Shape
- Predicting Car Prices Based on their Images
- Fundamentals of Convolutional Networks with LeNet5 on Kaggle's Digit Recognition Competition
- Dropout Visualization on MNIST Images
- Looking into LeNet5 with Random Weight Initialization
- Classifying Pet Images by Breed
- Intuitions about Image Embeddings and Pairwise Comparison
- Scraping Public Domain Artwork from Bing Images
- Extracting Embeddings with a Pretrained Resnet34
- Visualizing Embedding and Clustering Vector Spaces with FiftyOne
- Visualizing Image Embeddings with TensorBoard
- Running FiftyOne on a Github Codespace
- Image Neighborhoods with sklearn
- Running Imagenet-pretrained Resnets for classification
- Interpretability with Class Activation Mapping
- Fine-tuning a Resnet for multilabel classification
- Semantic Segmentation with U-Net
- Introduction to CLIP and Zero Shot Classification
- Zero-shot Object Detection with Grounding DINO
You can find us in the channel #practical-computer-vision-workshops to discuss the content
- FiftyOne community Discord - Invite Link
- Direct channel link (you must use the Invite Link first)
Feel free to fork and use the content, please keep the Creative Commons Attribution License.
antonio.rueda.toicen 'at' hpi 'dot' de
This work is licensed under a Creative Commons Attribution 4.0 International License.