There are 1 repository under multi-scale-features topic.
This project explores the impact of Multi-Scale CNNs on the classification of EEG signals in Brain-Computer Interface (BCI) systems. By comparing the performance of two models, EEGNet and MSTANN, the study demonstrates how richer temporal feature extractions can enhance CNN models in classifying EEG signals
[ICCV 2021] Official implementation of the paper "TRAR: Routing the Attention Spans in Transformers for Visual Question Answering"
Official implementation of "Exploiting Scale-Variant Attention for Segmenting Small Medical Objects"
The code of AGCN (Attention-driven Graph Clustering Network), which is accepted by ACM MM 2021.
An image classification neural network using multi scale features.