PriMAT: Performance of Lemur and Macaque Tracking Models

PriMAT tracking model performance data is available on figshare. The dataset, authored by Richard Vogg, was last updated on April 30, 2026. It contains evaluation results for a bounding-box-based model tested on Assamese macaques, redfronted lemurs, and other primate species.

Use Cases

• Benchmarking multi-animal tracking models based on bounding-box detection performance described in the study.
• Training individual identification models for primates based on the described classification branch achieving 84% accuracy for lemurs.
• Evaluating model transferability to other primate species like chimpanzees and gorillas as mentioned in the description.
• Studying the robustness of tracking in variable wild conditions like complex motion and occlusion, as outlined in the research.

Strengths

• Model achieved 84% accuracy in lemur identity prediction as stated in the description.
• Training required only a few hundred labeled frames, suggesting potential for efficient annotation.
• Evaluation includes transfer results to multiple species: Barbary macaques, Guinea baboons, chimpanzees, and gorillas.

Limitations

• Dataset is very small at 5.5 KB, indicating limited scope, likely containing only summary performance metrics.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment for statistical analysis.

Provenance

Source

Richard Vogg, ecker-lab (GitHub repository provided).

Collection Method

Results from applying the PriMAT tracking and identification model to video data of wild primates.

Freshness

Last updated 2026-04-30 17:45:17; freshness should be verified.

Geography

Field studies likely in habitats of Assamese macaques and redfronted lemurs; specific locations not stated.