Benchmarking Results
Computer Vision Model Performance
Explore the performance of our computer vision models tested across various behavioral paradigms. These results demonstrate the precision and accuracy of our advanced tracking and analysis systems.
Computer Vision Model Performance
| Maze | Pose Precision (%) | Pose mAP (%) | Precision (%) | Recall (%) | mAP50 (%) | Fitness Score |
|---|---|---|---|---|---|---|
|
Radial Arm / 8-Arm Maze |
96.98 |
97.56 |
97.0 |
99.5 |
99.5 |
1.795 |
|
Y Maze |
99.01 |
99.50 |
99.0 |
100.0 |
99.5 |
1.779 |
|
Water Maze / Morris Water Maze |
90.30 |
99.50 |
90.3 |
100.0 |
99.5 |
1.990 |
|
Novel Object Recognition |
98.95 |
99.27 |
99.0 |
99.0 |
99.3 |
1.754 |
|
Barnes Maze |
97.50 |
98.60 |
96.0 |
99.0 |
99.0 |
1.812 |
|
T Maze |
95.00 |
97.10 |
94.0 |
98.5 |
98.7 |
1.700 |
|
Elevated Plus Maze |
93.60 |
95.20 |
93.0 |
97.0 |
96.5 |
1.620 |
|
Sociability Chamber |
92.80 |
94.30 |
92.0 |
96.5 |
95.8 |
1.610 |
Pose Precision
Pose mAP
Precision & Recall
mAP50
Fitness Score
Explore our models in action!
Contact us to revolutionize your research!
Use Case Examples
Real-world applications where these metrics matter, such as:
Pose Precision
The accuracy of keypoint detection, indicating how precisely the model identifies landmarks.
mAP
The accuracy of keypoint detection, indicating how precisely the model identifies landmarks.
FAQ
What is Pose mAP?
Pose mAP (mean Average Precision) is a metric used to evaluate the performance of pose estimation models. It measures how accurately the model detects and localizes keypoints of a person or object in an image. A higher mAP value indicates better accuracy in pose detection.
How were these models tested?
The models were tested using standard benchmark datasets for object detection and pose estimation tasks. These datasets include annotated images with ground truth data, allowing the model’s predictions to be compared against the correct answers. Performance is then evaluated using metrics like Pose Precision, Pose mAP, Precision, Recall, and mAP50.
Can I use this for my project?
Yes! Depending on your project requirements, these models can be integrated into various applications, including healthcare, robotics, or behavioral studies. You can leverage the models for tasks like gait analysis, object tracking, or pose recognition. Contact us for more details about how to get started with these models for your specific use case.
