Application Problem Definition: Detecting human activities from still images

• Detecting human activities from still images remains a relatively challenging area for computer vision. Your task is to make use of a small portion of specifically selected images from MSCOCO, a large resource of labeled photos used to train and evaluate many models in computer vision. Your aim is to build a classifier that can successfully classify the specific images with the IDs provided in MyAberdeen according to which human action is taking place in the image (e.g. sitting, standing). The dataset includes multiple activity labels and multiple images. The image_ids are provided along with the image urls that you can download from the MSCOCO website (full instructions in template code provided).

Run project

• Create virtual environment

python -m venv venv

• Activate virtual environment

.\venv\Scripts\activate

• Install dependencies

pip install -r requirements.txt

Run the notebook