Deep learning based auto-labeling for sparse underwater freestyle marker-based optical motion capture using Qualisys Miqus M5U MoCap cameras
Bibliographic record
Abstract
This thesis proposes novel algorithms to automate cleaning and labeling of motion capture (MoCap) data specifically for underwater marker-based optical MoCap systems. Challenges related to sparse underwater freestyle MoCap data, captured using Qualisys Miqus M5U MoCap cameras, are explored using a dataset of 21 passive markers. A thorough review on MoCap denoising, recovery, alignment, and auto-labeling methods is conducted. The manual cleaning process using Qualisys Track Manager software and Automatic Identification of Markers function is explained. Then, a novel semi-supervised geometry-based labeling algorithm is developed based on distance and angle measurements with a visual evaluation of 100% accuracy. This algorithm includes sub algorithms for extraneous removal via norm differences, anomaly detection, pelvis detection based on Principal Component Analysis, recovery of missing markers, and detection of corresponding reappearing markers along with a side detection algorithm. Finally, a deep learning-based auto-labeling algorithm utilizing Long-Short-Term Memory is proposed, employing Hungarian label assignment and Procrustes analysis to label unlabeled data. The network accepts the 3D relative positions of markers, velocity, and acceleration. The ground truth and the training set are generated by the geometry-based algorithm and enhanced using data augmentation and transfer learning of simulated trajectories. The pelvis detection technique automates the alignment, and the extraneous removal algorithm enhances accuracy from 66% to 98%. These algorithms work effectively in the presence of outliers, extraneous, ghosts, and missing markers. Future work will evaluate the algorithm with more data and ghost markers and explore a more robust body side detection algorithm.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.002 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".