WildLife MegaDetector

Our new model ZOO works with DepthAI V3. Find out more in our documentation.

1 Likes

Model Details

Model Description

WildLife MegaDetector model is a YOLO based model specifically designed for animal detection. Besides animals model model detects humans and vehicles allowing us to monitor wild life, farms, and pastures (3 classes: animal, person, vehicle). The model is designed for computational efficiency and performance.

Developed by: Microsoft

Shared by:

Model type: Computer Vision

License:

Resources for more information:

Training Details

Training Data

The training dataset is not publicly accessible.

Testing Details

Metrics

Results are taken from .

Model	Params ^(M)	mAR (Animal class)	mAP@50 (all classes)
MegaDetectorV6-YOLOv10-Compact	2.3	76.8	87.2

Note: Animal Recall is included as a primary performance metric, even though it is not commonly used in traditional object detection studies, which typically focus on balancing overall model performance. For MegaDetector, the goal is to optimize for animal recall—in other words, minimizing false negative detections of animals or, more simply, ensuring the model misses as few animals as possible. While this may result in a higher false positive rate, authors rely on downstream classification models to further filter the detected objects. Authors believe this approach is more practical for real-world animal monitoring scenarios.

Technical Specifications

Input/Output Details

Input:

Name: images

Info: NCHW BGR un-normalized image

Output:

Name: multiple (see NN archive)

Info: Unprocessed outputs of a multitude of detections

Model Architecture

Backbone: an enhanced version of CSPNet (Cross Stage Partial Network)

Neck: includes PAN (Path Aggregation Network)

Head: One-to-One head

Please consult the for more information on model architecture.

Throughput

Model variant: wildlife-megadetector:mdv6-yolov10-c

• Input shape: [1, 3, 288, 512] • Output shapes: [[1, 8, 36, 64], [1, 8, 18, 32], [1, 8, 9, 16]]

• Params (M): 2.266 • GFLOPs: 1.392

Platform	Precision	Throughput (infs/sec)	Power Consumption (W)
RVC2	FP16	29.87	N/A
RVC4	FP16	604.10	3.88

* Benchmarked with , using 2 threads (and the DSP runtime in balanced mode for RVC4).

* Parameters and FLOPs are obtained from the package.

Utilization

Models converted for RVC Platforms can be used for inference on OAK devices. DepthAI pipelines are used to define the information flow linking the device, inference model, and the output parser (as defined in model head(s)). Below, we present the most crucial utilization steps for the particular model. Please consult the docs for more information.

Install DAIv3 library:

pip install depthai

Define model:

model_description = dai.NNModelDescription(
    "luxonis/wildlife-megadetector:mdv6-yolov10-c"
)

nn = pipeline.create(dai.node.DetectionNetwork).build(
    <CameraNode>, model_description
)

The model is automatically parsed by DAI and it outputs the message (bounding boxes, labels, and scores of the detected objects).

Get model output(s):

while pipeline.isRuning():
    nn_output: dai.ImgDetections = parser_output_queue.get()

Example

You can quickly run the model using our script. It automatically downloads the model, creates a DepthAI pipeline, runs the inference, and displays the results using our DepthAI visualizer tool. To try it out, run:

python3 main.py \
    --model luxonis/wildlife-megadetector:mdv6-yolov10-c

YOLOv10 based animal detector.
License	GNU Affero General Public License v3.0 Commercial use
Downloads	2044
Tasks	Object Detection
Model Types	ONNX PYTORCH

Model Variants

Name	Version	Available For	Created At	Deploy
		RVC2, RVC4	11 months ago