Use Case and High-Level Description

This is the UNet model that is designed to perform semantic segmentation. The model has been trained on the CamVid dataset from scratch using PyTorch framework. Training used median frequency balancing for class weighing. For details about the original floating point model, check out the paper.

The model input is a blob that consists of a single image of "1x3x368x480" in BGR order. The pixel values are integers in the [0, 255] range.

The model output for unet-camvid-onnx-0001 is the per-pixel probabilities of each input pixel belonging to one of the 12 classes of the CamVid dataset.


Metric Value
GFlops 260.1
MParams 31.03
Source framework PyTorch*


The quality metrics were calculated on the CamVid validation dataset. The 'unlabeled' class had been ignored during metrics calculation.

Metric Value
mIoU 71.95%



Image, shape - 1,3,368,480, format is B,C,H,W where:

Channel order is BGR


Semantic segmentation class probabilities map, shape -1,12,368,480, output data format is B,C,H,W where:

Legal Information

[*] Other names and brands may be claimed as the property of others.