AI and ML Workflows Using Backblaze B2

These Backblaze B2 Cloud Storage sample projects demonstrate how to build practical AI- and data-driven workflows that integrate seamlessly with Backblaze B2, ranging from fully client-side inference in the browser to server-side machine-learning pipelines.

Data Orchestration

This notebook shows how to use Pixeltable to extract and manage video frames stored in Backblaze B2, organizing the frames into a queryable, multimodal dataset that can support downstream AI and data processing workflows such as indexing, transformation, and model inference.

For basic instructions about integrating PixelTable with Backblaze B2, click here.

Model Training

This Jupyter notebook demonstrates training and evaluating an image classification model on the CIFAR-10 dataset using PyTorch, showing how to load, preprocess, train, and test a neural network for computer vision tasks with dataset and artifacts stored or tracked in conjunction with Backblaze B2.

Client-side AI Inference

Audio

This sample demonstrates a browser-based speech-to-text transcription app using Transformers.js and the Whisper model, allowing users to transcribe audio locally in the browser and upload both the original audio and generated transcript to Backblaze B2.

Images

This sample demonstrates a browser-based image background removal application using Transformers.js and the RMBG-1.4 model, enabling users to remove backgrounds locally in the browser and store both the original and processed images in Backblaze B2.