Naoki Shibuya

This is a blog article about understanding the upsampling with transposed convolution works for those who might have been confused about the logic. This is one of the posts I wrote for Medium, and it was the most popular one.

Built and trained a convolutional network for end-to-end driving in a simulator using Tensorflow and Keras. I implemented the driving (steering and throttle) model based on the Nvidia paper on deep learning and self-driving cars and using Udacity's car simulation environment for training and testing.

Link to the Nvidia paper used for this project:
https://devblogs.nvidia.com/parallelforall/deep-learning-self-driving-cars/

I've written different deep learning examples to accompany some of my Medium blog articles (http://naokishibuya.medium.com). I covered the topics such as style transfer, autoencoder, generative adversarial network (GAN), and transposed convolution.

This is the capstone project I did for Udacity's Cloud Developer Nanodegree in which I used AWS CDK (TypeScript) to allocate resources for AWS API Gateway, Lambda, S3, and DynamoDB and use them for the simple to-do list manager. I also used React, Redux, and Redux Thunk for the front end.

I've written a series of articles on transformer architecture:
• Encoder-decoder architecture
• Self-attention mechanism
• Positional encoding
• Word embedding
• Beam search
• BLEU
I've also drawn technical diagrams to explain the details of the architecture.