mmaaz60/EdgeNeXt
A hybrid CNN-Transformer architecture for efficient image classification on mobile and edge devices.
Not currently ranked — collecting fresh signals.
star history
EdgeNeXt provides a efficient amalgamated architecture combining convolutional neural networks with transformer blocks for mobile vision tasks. The model is designed for edge deployment, achieving strong ImageNet classification accuracy with low latency suitable for mobile devices. The repository includes pretrained weights and training scripts for the classification task.
Frequently asked
- What is mmaaz60/EdgeNeXt?
- A hybrid CNN-Transformer architecture for efficient image classification on mobile and edge devices.
- Is EdgeNeXt open source?
- Yes — mmaaz60/EdgeNeXt is open source, released under the MIT license.
- What language is EdgeNeXt written in?
- mmaaz60/EdgeNeXt is primarily written in Python.
- How popular is EdgeNeXt?
- mmaaz60/EdgeNeXt has 417 stars on GitHub.
- Where can I find EdgeNeXt?
- mmaaz60/EdgeNeXt is on GitHub at https://github.com/mmaaz60/EdgeNeXt.