This repository explores and implements concepts from the research paper "Bottleneck Transformers for Visual Recognition" by Aravind Srinivas, Tsung-Yi Lin, and others from Google Research. The repository contains both a detailed presentation of the study and Jupyter notebooks with practical implementations.
- Title: Bottleneck Transformers for Visual Recognition
- Authors: Aravind Srinivas, Tsung-Yi Lin, et al.
- Link: Read the paper
- Published: 2021
Bottleneck Transformers.pdf: A presentation that outlines the key concepts and findings from the paper.Bottleneck_Transformer.ipynb: Jupyter notebook with the implementation of the bottleneck transformer model as described in the paper.
The research introduces a novel hybrid architecture that combines the robustness of CNNs with the efficiency of self-attention mechanisms from transformers to address the challenges in visual recognition tasks such as object detection and image classification.
Please run our notebook Bottleneck_Transformer.ipynb top to bottom
Will Qi, Tin Do, Tarunyaa Sivakumar