Skip to content

harvard-edge/Wake_Vision_Quickstart

Repository files navigation

Wake Vision Logo

Wake Vision Quickstart

WakeVision.ai

What is Wake Vision?

"Wake Vision" is a large, high-quality dataset featuring over 6 million images, significantly exceeding the scale and diversity of current tinyML datasets (100x). This dataset includes images with annotations of whether each image contains a person. Additionally, it incorporates a comprehensive fine-grained benchmark to assess fairness and robustness, covering perceived gender, perceived age, subject distance, lighting conditions, and depictions. Annotations are published under a CC BY 4.0 license, and all images are sourced from the Open Images v7 dataset under a CC BY 2.0 license.

Colab

Colab Link

We provide a Colab to quickly interact with the dataset without downloading anything.

Quickstart

Install Requirements

pip install -r requirements.txt

Run Benchmark

The dataset will automatically download if you use HuggingFace Datasets. Currently, you need to build the TFDS version of the dataset manually.

python3 benchmark.py -m"example_wake_vision_mobilenetv2.keras" -d=$DATASET_DIR -t={"tfds", "hf"}

Train a MobileNetV2_0.25 on Wake Vision

python3 train.py -m="Wake_Vision_MNV2.keras" -d=$DATASET_DIR -t={"tfds", "hf"}

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages