🌐 How should we standardize data organization in markerless motion capture? 📊
In markerless motion capture, effective data organization is crucial for ensuring data is shareable, comparable, and easy to process. But how should this be done?
👉 The Big Question:
Should we aim for a broad standardization for all markerless setups? Or should we focus on specific use cases, like comparing different methods?
We believe that two elements are critical for standardization:
1️⃣ File Types: Defining which formats (e.g., .avi, .hdf5, .toml) should be used for each data type.
2️⃣ Metadata Requirements: Establishing a minimal set of metadata for each dataset.
3️⃣ Data Structure: Determining how data should be organized across folders for flexibility and scalability.
📁 Proposed Structures:
By Data Type: Each folder can be processed independently, making it easier to share specific datasets.
By Trial: A complete trial with all associated data in one folder, representing the real-world acquisition process.
💡 We want your input! How would you organize data for markerless motion capture?
Would you prefer a general standard for all setups or separate standards for specific purposes?
What are the most critical metadata fields you would include?
Do you prefer a structure by data type or by trial?
Share your thoughts below! 💬 Let’s discuss and collectively design a robust data standard for the community. #Biomechanics #MarkerlessMotionCapture #DataOrganization
🌐 How should we standardize data organization in markerless motion capture? 📊
In markerless motion capture, effective data organization is crucial for ensuring data is shareable, comparable, and easy to process. But how should this be done?
👉 The Big Question:
Should we aim for a broad standardization for all markerless setups? Or should we focus on specific use cases, like comparing different methods?
We believe that two elements are critical for standardization:
1️⃣ File Types: Defining which formats (e.g., .avi, .hdf5, .toml) should be used for each data type.
2️⃣ Metadata Requirements: Establishing a minimal set of metadata for each dataset.
3️⃣ Data Structure: Determining how data should be organized across folders for flexibility and scalability.
📁 Proposed Structures:
By Data Type: Each folder can be processed independently, making it easier to share specific datasets.
By Trial: A complete trial with all associated data in one folder, representing the real-world acquisition process.
💡 We want your input! How would you organize data for markerless motion capture?
Would you prefer a general standard for all setups or separate standards for specific purposes?
What are the most critical metadata fields you would include?
Do you prefer a structure by data type or by trial?
Share your thoughts below! 💬 Let’s discuss and collectively design a robust data standard for the community. #Biomechanics #MarkerlessMotionCapture #DataOrganization