The Model Garden by Neuralet is a directory of the most common pre-trained deep learning models that are deployable on edge devices, such as the NVIDIA Jetson and Coral Edge TPU platforms. These models are categorized by task and device to help you choose the one that matches your needs at a glance. Performance measures, including throughput (FPS) and inference time (ms), are also listed for each model to allow comparison between platforms.
Step 1: Choose the device
To explore the Model Garden, first, choose the device you want to work with. Here is a list of currently-supported platforms:
We will keep adding more devices to the Model Garden.
Step 2: Choose the task
We have collected pre-trained models for the most common tasks in computer vision:
- Image Classification
- Object Detection
- Pose Estimation
- Semantic Segmentation
Choose the desired task to view the list of available models.
Step 3: EXPLORE!
Compare inference time (ms) and throughput (FPS) of the listed models, read the descriptions to learn more about the specifics, and visit the URLs to decide which model you want to use. Explore the Model Zoo, and click the model download link to start using the model.
Feel free to report model performances, give comments, answer others, and ask new questions about each model on the model’s page. We are happy to help!