Deep learning-augmented super-resolution microscopy reconstruction
2019.11 - Present, Postdoc & Ph.D., Institut Pasteur, Paris
The objective of this project is to improve the quality of super-resolution images through the deep learning techniques, with the primary aim of minimizing the number of acquired single-molecule images. Accelerating the acquisition process has the potential to facilitate high-resolution visualization of biological structures within living cells, with a particular focus on dynamic microtubules.
Achievements:
- Developed and implemented a deep learning algorithm for image/video translation using Python and TensorFlow.
- Executed scripts on remote GPU/CPU clusters to optimize computational performance.
- Pioneered the application of the vision transformer to leverage temporal information for enhanced results.
- Published one paper, with another paper currently in the revision stage and a third paper in the preparation phase.