Introduction
Adilkhan and Darkhan are young researchers who decided to explore how technology can improve physical training.
They equipped participants with IMU sensors and recorded multiple yoga pose repetitions.
Each attempt was carefully labeled as either correct or incorrect.
Their goal is simple: to train a model that can automatically recognize whether a repetition is done properly.
Now, it is your mission to continue their work and build an accurate classifier to evaluate unseen yoga repetitions.
Formal Problem Statement
You are given three CSV files:
X_train.csv: Training set with IMU data segments.y_train.csv: Labels for the training set (0= incorrect,1= correct).X_test.csv: Test set with IMU data segments, without labels. Your model must predict the label for each test example.
Dataset links:
- X_train.csv: https://drive.google.com/file/d/1wMuHEYsXKt3_IhgAUiUcpPvuctLM9jGL/view?usp=drive_link
- y_train.csv: https://drive.google.com/file/d/1F3mUFseTH3mi6bkY13flJTAK4tGTMIzm/view?usp=drive_link
- X_test.csv: https://drive.google.com/file/d/1OdlLTsLPyLa7JUmhQcRtP9ixlYE82H7k/view?usp=drive_link
Input
Features (X)
Each repetition is represented by time series of IMU readings sampled at 200 Hz. The following sensor signals are provided:
ax,ay,az— Linear acceleration (g) along the sensor’s X, Y, Z axes.wx,wy,wz— Angular velocity (deg/s) around the sensor’s X, Y, Z axes.
Labels (y)
0— repetition performed incorrectly.1— repetition performed correctly.
Each id corresponds to one yoga repetition.
Output
Participants must submit only a CSV file named solution.csv, with the following format:
id,label
0,1
1,0
2,1
...
Important: Submit only the solution.csv file. Source code is not required. The order of id values in solution.csv must strictly follow the order in test.csv, starting from 0 up to the highest ID.
Evaluation
- Your predictions will be evaluated using the Accuracy score.
Example
Input (X_test.csv):
id,ax,ay,az,wx,wy,wz
0,0.12,0.04,0.98,-1.2,0.3,0.1
1,0.01,0.15,1.02,0.8,-0.5,0.0
2,-0.10,0.20,0.95,0.7,0.1,-0.2
Your Output (solution.csv):
id,label
0,1
1,0
2,1