Cos-Triplet Ensemble Learning Model for Quality Scoring and Classification on Rehabilitation Pose Matching System
Bibliographic record
Abstract
In a pose matching and guidance system pertinent to rehabilitation training, the establishment of a distinct decision boundary demarcating qualified and unqualified poses is of utmost importance, particularly in domains necessitating precise pose alignment, such as rehabilitation and martial arts training. Moreover, it is imperative for patients to receive real-time feedback regarding their postural alignments during the training sessions. Accomplishing these objectives with high precision and performance presents a formidable challenge. This paper introduces a Cos-Triplet Ensemble Network Combined with Classification Module (CTENM) model, meticulously designed to furnish an accurate pose-matching score concomitant with exceptional precision and accuracy in pose classification. The CTENM model is an intricate stacking model comprising three sub-models, each of which is a costriplet network fortified with varying CNN backbones. In our experiment, VGG16, MobileNet, and ResNet18 were judiciously selected as the backbones for these sub-models. The cos-triplet model represents an innovative modification of the original triplet network. This adaptation incorporates cosine similarity triplet loss as a substitute for the conventional triplet contrastive loss, further augmented with a stochastic weighting mechanism to facilitate classification in triplet settings. Post-training of all sub-models, the stacking ensemble model endeavors to enhance its capabilities by aggregating weighted outputs from the three sub-models. These weight parameters are subsequently refined through training a more comprehensive cos-triplet network. The empirical results gleaned from our experiments affirm the efficacy of the CTENM model in accomplishing the dual tasks of pose matching and guidance with unparalleled accuracy and precision.
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) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| 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".