🦴📊 Skeletons Don’t Lie: Can AI Decode Your Moves? 🤖

🧾 Description

We are given a dataset containing sequences of human movements, each composed of multiple frames (FrameNumber).
Each frame includes the 3D positions of 25 body joints, represented by the coordinates:
J1X, J1Y, J1Z, ..., J25X, J25Y, J25Z.

Each frame is associated with a sequence identifier (IDSample) and two labels:

• Action — the action performed
• Camera — the camera that recorded the sequence

To better understand the spatial distribution of the 25 joints:

🔢 Possible Actions

• 0: wear jacket
• 1: take off a hat/cap
• 2: hopping (one foot jumping)
• 3: jump up
• 4: falling

🎥 Possible Cameras

• 1: 45° angle view (left or right)
• 2: front view
• 3: side view (left or right)

🧮 Subtask 1 — Counting Frames per Sequence (10 points)

For each unique IDSample, determine how many frames are available in the test dataset (test_data.csv).
This analysis helps understand the temporal-spatial distribution of the data for each action.

🧠 Subtask 2 — Sequence-Level Action Classification (40 points)

Using the training data, train a classification model capable of recognizing the action performed in a sequence of frames.
Apply the model to the test data and, for each unique IDSample, predict the most probable action.

📊 Evaluation Metric

The performance of the model will be evaluated using accuracy, defined as:

To obtain full points, the achieved accuracy must be at least 0.965.

🎥 Subtask 3 — Camera Identification (50 points)

Using the training data, train a model capable of predicting which camera recorded a given sequence of frames.
Apply the model to the test set and specify, for each IDSample, the camera predicted as most likely.

📊 Evaluation Metric

The model performance is again evaluated using accuracy, defined as:

To obtain full points, the achieved accuracy must be at least 0.8.

📦 Output Format

The final result must be a CSV file named output.csv, containing exactly 3 columns:

🗂️ Useful Resources

• Complete Starter Kit – includes a skeleton to start solving the problem.
• Problem Statement PDF – contains the detailed description of the task and figures.