SOLO - 3.5 English

Vitis AI Library User Guide (UG1354)

Document ID
Release Date
3.5 English

Segment objects by locations (SOLO) is a simple and flexible framework applied for accomplishing instance segmentation in digital image processing and computer vision tasks. It is based on the notion of “instance categories” for instance segmentation in which each pixel within an instance of an object is assigned a category based on its location and size.

The following table lists the SOLO model supported by the Vitis AI Library.

Table 1. SOLO Models List
No Model Name Framework
1 solo_pt PyTorch