Extract Visual Features Tools

Fashion-Keras

Development

The industry-standard drop-in replacement for MNIST for benchmarking fashion-centric deep learning models.

OpenCLIP

Computer Vision

The industry-standard open-source implementation of Contrastive Language-Image Pre-training (CLIP).

Swin Transformer

Artificial Intelligence

Hierarchical Vision Transformer using Shifted Windows for general-purpose computer vision tasks.

Vision Transformer (ViT) Large

General AI

A large-sized Vision Transformer model pre-trained on ImageNet for image classification tasks.

HRNet-Semantic-Segmentation

General AI

High-resolution networks for semantic segmentation tasks.

Places365

Computer Vision

The industry-standard deep learning dataset and model suite for state-of-the-art scene recognition.

FFD (Fast Feature Detection) AI

Development

Sub-millisecond computer vision feature extraction for edge-native AI applications.