Figures and data
Experimental setup of the camera system and primate chair for face detection and identification.
(A) Side view of the camera (green box) and primate chair, attached to the housing cage. (B) Frontal view of the camera fixed on the primate chair, enclosed within a protective cover. (C) A schematic illustration of the real-time facial image recording and automatic identification of the marmoset entering the primate chair.
Workflow and design of the marmoset facial detection and identification model.
(A) The architecture of the real-time marmoset facial recognition program. (B) Marmoset face images from three camera angles. (C) Bounding boxes of marmoset faces (green box) and collars (pink box) were manually labeled in the training and validation datasets to train the multi-marmoset face classification model. (D) The schematic of the automatic face (blue box) and collar bead (cyan box) extraction model.
Performance comparison of multi-marmoset face classification models.
Comparison of Recall, Precision, F1 score, mAP at IoU = 0.5:0.95, Validation DFL, and training time of the three pre-trained model, based on the performance of marmoset face detection and identification on the adult marmosets’ dataset. The highest values of each parameter were highlighted in bold.
Training performance of the multi-marmoset face classification model for three adult marmosets, using the pre-trained YOLOv8 nano model.
(A) Precision of all detection classes across training epochs. The red dotted line denoted the final model with the best performance at training epoch 183. (B) Similar to (A), except for overall recall. (C) Similar to (A), except for the mAP at the IoU at 0.5:0.95. (D) Overall Model precision, recall, and the mAP@50-95 (IoU = 0.5:0.95) for each label class.
Normalized confusion matrix per-class classification across the 6 label classes in the adult marmoset recognition model.
The y-axis represents the predicted class, and the x-axis represents the manually labelled class. Proportion was generated by the (A) training dataset and the (B) validation dataset, showing whether certain classes were frequently mislabelled as a different class.
Training performance of the automatic facial and identity extraction model.
(A) Precision of all detection classes across training epochs. The red dotted line denoted the final model with the best performance at training epoch 124. (B) Similar to (A), except for overall recall. (C) Similar to (A), except for the mAP at the IoU at 0.5:0.95. (D) Overall Model precision, recall, and the mAP@50-95 (IoU = 0.5:0.95) for each label class.
Normalized confusion matrix per-class classification across the 2 label classes in the automatic identity extraction model.
Proportion was generated by the (A) training dataset and the (B) validation dataset, showing whether certain classes were frequently mislabelled as a different class.
Training performance of the multi-marmoset face recognition model for young marmosets at 7 months.
(A) Precision of all detection classes across training epochs. The red dotted line denoted the final model with the best performance at training epoch 245. (B) Similar to (A), except for overall recall. (C) Similar to (A), except for the mAP at the IoU at 0.5:0.95. (D) Overall Model precision, recall, and the mAP (IoU = 0.5:0.95) for each label class.
Normalized confusion matrix per-class classification across the 4 label classes in the young marmoset recognition model.
Proportion was generated by the (A) training dataset and the (B) validation dataset, showing whether certain classes were frequently mislabelled as a different class.
Across-model visualization of the face similarity between marmoset pairs.
Four types of family relationships (mother-father, father-son, mother-son, and twin1-twin2) were compared, based on the training results of adult and young marmosets. The similarity was calculated using (A) cosine similarity and the (B) Euclidean distance. Heat map of the (C) cosine similarity and (D) Euclidean distance was plotted between the 3 relationship pairs in the adult family. The cosine similarity score ranged from 0 (very different) and 1 (exactly same) for marmoset faces. The Euclidean distance score ranged from 0 (exactly same) and 1 (very different) for marmoset faces.
Statistical analysis of inter-individual face similarity (cosine similarity) between different family relationships, within the adult marmoset family.
Comparison of t-statistics, p-value, and Cohen’s d on cosine similarities were compared between the 3 family relationships. The relationship pairs that showed significant differences were bolded.
Statistical analysis of inter-individual face similarity (Euclidean distance) between different family relationships, within the adult marmoset family.
Comparison of t-statistics, p-value, and Cohen’s d on Euclidean distances were compared between the 3 family relationships. The relationship pairs that showed significant differences were bolded.
Examples of marmoset face detection and classifier for in-cage and open-field marmoset images.
This image illustrated the (A, B, D, E) detection of marmosets faces and (C) their identities across different cameras, environments, and marmoset behaviors. Additionally, it included scenarios where animals were housed in-cage (A – C) or moving freely (D – E). The open-field marmoset videos (D - E) were obtained from the Youtube Billabong Zoo, Koala & Wildlife Park: https://www.youtube.com/watch?v=lDJNk4rjqqQ.
Identification performance of human experimenter across the 5 marmosets.
Identification accuracy of 7 human experimenters across the 5 marmosets. The legend denoted the performance of the animal health technicians (blue) and the student/supervisor (pink).
Training performance of the multi-marmoset face classification model for three adult marmosets.
The Validation Distribution Focal Loss (DFL) of all detection classes across training epochs. The red dotted line denoted the final model with the best performance at training epoch 183.
Training performance of the automatic facial and identity extraction model.
The Validation Distribution Focal Loss (DFL) of all detection classes across training epochs. The red dotted line denoted the final model with the best performance at training epoch 124.
Training performance of the multi-marmoset face classification model for two young marmosets.
The Validation Distribution Focal Loss (DFL) of all detection classes across training epochs. The red dotted line denoted the final model with the best performance at training epoch 245.
Heat map of the cosine similarity between marmoset faces in the young family.
The two young marmosets (twins) were represented by their family, with the similarity score calculated between the two twins. The cosine similarity score ranged from 0 (very different) and 1 (exactly same) for marmoset faces.
Heat map of the Euclidean distance between marmoset faces in the young family.
The two young marmosets (twins) were represented by their family, with the similarity score calculated between the two twins. The Euclidean distance score ranged from 0 (exactly same) and 1 (very different) for marmoset faces.
Model performance evaluation of the validation dataset in the adult marmoset recognition model.
Comparison of precision, recall, mAP@50, and mAP@50-95 were computed for all the detected label classes in this model.
Model performance evaluation of the validation dataset in the automatic face and identity extraction model.
Comparison of precision, recall, mAP@50, and mAP@50-95 were computed for all the detected label classes in this model.
Model performance evaluation of the validation dataset in the young marmoset recognition model.
Comparison of precision, recall, mAP@50, and mAP@50-95 were computed for all the detected label classes in this model.
