MATLAB Deep Learning Model Hub

Discover pretrained models for deep learning in MATLAB.

Models <a name="Models"/>

Computer Vision

• Image Classification
• Object Detection
• Semantic Segmentation
• Instance Segmentation
• Image Translation
• Pose Estimation
• 3D Reconstruction
• Video Classification
• Text Detection & Recognition

Natural Language Processing

• Transformers

Audio

• Audio Embeddings
• Sound Classification
• Pitch Estimation
• Speech to Text

Lidar

• Point Cloud Classification
• Point Cloud Segmentation
• Point Cloud Object Detection

Robotics

• Manipulator Motion Planning
• Path Planning with Motion Planning Networks

Image Classification <a name="ImageClassification"/>

Pretrained image classification networks have already learned to extract powerful and informative features from natural images. Use them as a starting point to learn a new task using transfer learning.

Inputs are RGB images, the output is the predicted label and score:

These networks have been trained on more than a million images and can classify images into 1000 object categories.

Models available in MATLAB:

Note 1: Since R2024a, please use the imagePretrainedNetwork function instead and specify the pretrained model. For example, use the following code to access googlenet:

[net, classes] = imagePretrainedNetwork("googlenet");

| Network | Size (MB) | Classes | Accuracy % | Location | | ------------- | ------------- | ------------- | ------------- | ------------- | | googlenet<sup>1<sup> | 27| 1000| 66.25 |Doc <br />GitHub | | squeezenet<sup>1<sup> | 5.2| 1000| 55.16 |Doc | | alexnet<sup>1<sup> | 227| 1000|54.10|Doc | | resnet18<sup>1<sup> | 44| 1000|69.49|Doc <br />GitHub | | resnet50<sup>1<sup> | 96| 1000|74.46|Doc <br />GitHub | | resnet101<sup>1<sup> | 167| 1000|75.96|Doc <br />GitHub | | mobilenetv2<sup>1<sup> | 13| 1000| 70.44|Doc <br />GitHub | | vgg16<sup>1<sup> | 515| 1000|70.29|Doc | | vgg19<sup>1<sup> | 535| 1000|70.42|Doc | | inceptionv3<sup>1<sup> | 89| 1000|77.07|Doc | | inceptionresnetv2<sup>1<sup> | 209| 1000|79.62|Doc | | xception<sup>1<sup> | 85| 1000|78.20|Doc | | darknet19<sup>1<sup> | 78| 1000|74.00|Doc | | darknet53<sup>1<sup> | 155| 1000|76.46|Doc | | densenet201<sup>1<sup> | 77| 1000| 75.85|Doc | | shufflenet<sup>1<sup> | 5.4| 1000|63.73|Doc | | nasnetmobile<sup>1<sup> | 20| 1000|73.41|Doc | | nasnetlarge<sup>1<sup> | 332| 1000|81.83|Doc | | efficientnetb0<sup>1<sup> | 20| 1000|74.72|Doc | | ConvMixer | 7.7| 10|-|GitHub | | Vison Transformer | Large-16 - 1100<br /> Base-16 - 331.4<br /> Small-16 - 84.7<br /> Tiny-16 - 22.2| 1000|Large-16 - 85.59<br /> Base-16 - 85.49<br /> Small-16 - 83.73<br /> Tiny-16 - 78.22|Doc |

Tips for selecting a model

Pretrained networks have different characteristics that matter when choosing a network to apply to your problem. The most important characteristics are network accuracy, speed, and size. Choosing a network is generally a tradeoff between these characteristics. The following figure highlights these tradeoffs:

Figure. Comparing image classification model accuracy, speed and size.

Back to top

Object Detection <a name="ObjectDetection"/>

Object detection is a computer vision technique used for locating instances of objects in images or videos. When humans look at images or video, we can recognize and locate objects of interest within a matter of moments. The goal of object detection is to replicate this intelligence using a computer.

Inputs are RGB images, the output is the predicted label, bounding box and score:

These networks have been trained to detect 80 objects classes from the COCO dataset. These models are suitable for training a custom object detector using transfer learning.

| Network | Network variants | Size (MB) | Mean Average Precision (mAP) |Object Classes | Location | | ------------- | ------------- |------------ | ------------- |------------ |------------ | | EfficientDet-D0 | efficientnet |15.9 | 33.7 |80 |GitHub | | YOLO v9 | yolo9t<br />yolo9s<br />yolo9m<br />yolo9c<br />yolo9e | 7.5 <br /> 25 <br /> 67.2 <br /> 85 <br />190 | 38.3<br /> 46.8<br /> 51.4<br />53.0 <br />55.6|80 |GitHub| | YOLO v8 | yolo8n<br />yolo8s<br />yolo8m<br />yolo8l<br />yolo8x | 10.7 <br /> 37.2<br />85.4 <br />143.3<