Skip to main content
Visual Computing for Industry, Biomedicine, and Art

Table 2 Our adapted Transformer parameters used in the encoder

From: Multi-task approach based on combined CNN-transformer for efficient segmentation and classification of breast tumors in ultrasound images

ViT version

Image resolution

Projection dimension

Number of MSA heads

Number of transformers layers

ViT-base

224 × 224

768

12

12

Our adapted transformer

256 × 256

64

8

12

Back to article page