**Participants will:**

Develop state-of-the-art models for elderly action recognition using publicly available datasets.
Showcase innovative techniques for data curation and transfer learning.
Contribute to a growing body of research that addresses real-world challenges in elderly care.

**Key Dates**
Challenge Launch: January 10, 2025
Submission Deadline: February 15, 2025

**Winners Announcement:**
February 20, 2025, at the virtual AI, Machine Learning, and Computer Vision Meetup

**Model Development**
Participants are encouraged to explore and leverage state-of-the-art human action recognition models without limitations. They are also strongly encouraged to be creative and original in their model architecture and training methodology.

**Dataset**
To mitigate overfitting and promote generalization, participants could build training datasets using publicly available resources such as:

[ETRI-Activity3D](https://etri.org/Activity3D)
[MUVIM](https://muvim.org/)
[ToyotaSmartHome](https://toyotasmarthome.com/)

Just so you know, some datasets require submitting a request before downloading. Participants are not restricted to these datasets and are welcome to curate extensive datasets, combining multiple sources. A detailed report outlining the datasets used and their preparation and curation processes is mandatory.

**Data Curation and Categorization**
For efficient organization and analysis, participants must group activities into the following categories. The model output should display both categories and activities.

- Locomotion and Posture Transitions include walking, sitting down/standing up, getting up/lying down, exercising, and looking for something.
- Object Manipulation: Spreading bedding/folding bedding, wiping table, cleaning dishes, cooking, vacuuming
- Hygiene and Personal Care: Washing hands, brushing teeth, taking medicine
- Eating and Drinking
- Communication and Gestures: Talking, phone calls, waving a hand, shaking hands, hugging
- Leisure and Stationary Actions: Reading, watching TV

Keep in mind that the submission file should have the following categories, all lowercase: ```locomotion, manipulation, hygiene, eating, communication, and leisure.```

Supporting Materials
Voxel51’s [FiftyOne](https://voxel51.com/fiftyone/) tool will assist participants in curating, categorizing, and visualizing datasets effectively. The challenge launch will provide tutorials and examples.

Find some blogs and notebooks here: 
https://medium.com/@paularamos_phd/journey-into-visual-ai-exploring-fiftyone-together-part-iii-preparing-a-computer-vision-e5709684ee34 
https://medium.com/@paularamos_phd/journey-into-visual-ai-exploring-fiftyone-together-part-iv-model-evaluation-8a83ccdbe021