WhatDoWeSee / README.md
StephanMeijer's picture
EfficientNetV2-B0 v0.4.0 as default; v0.4.0-rc.1 out of the menu
803273b verified
|
Raw History Blame Contribute Delete
1.11 kB
metadata
title: WhatDoWeSee
emoji: 🏷️
colorFrom: indigo
colorTo: blue
sdk: static
pinned: false
license: apache-2.0
short_description: Tag what is visible in an image, in your browser
models:
  - StephanMeijer/WhatDoWeSee-EfficientNetV2-B0
  - StephanMeijer/WhatDoWeSee-EfficientNet-B0
  - StephanMeijer/WhatDoWeSee-MobileNetV4

WhatDoWeSee

Drop an image and a small model tags what is visible in it: people, animals, text, charts, diagrams, tables, maps, aerial views and more. Everything runs in your browser with ONNX Runtime Web (WebAssembly): the image is never uploaded.

The models are loaded from StephanMeijer/WhatDoWeSee-EfficientNetV2-B0 (the default), StephanMeijer/WhatDoWeSee-MobileNetV4 and StephanMeijer/WhatDoWeSee-EfficientNet-B0, each at a fixed version. The browser resizes the image itself, so scores can differ slightly from the Python reference implementation in the model card.