lone15
@lone15
Profile
Registered: 1 year, 4 months ago
How did the AI painting technology that went viral in the community develop? Keras and TensorFlow have a very deep relationship. Keras is Python's deep learning API, built on top of TensorFlow, and provides an easy way to define and train deep learning models. Since the end of 2021, Keras has been used by more than one million people, including engineers, data scientists and academic researchers in major companies and startups. Even Youtube's video recommendation system and Waymo's self-driving car are developed using Keras, and it is a popular framework on Kaggle. Keras appeared 8 months earlier than TensorFlow. At the beginning, it was built on Theano (a library developed by the Montreal Institute of Learning Algorithms at the University of Montreal). After the release of TensorFlow, users can switch between the two . With the maturity of TensorFlow technology and the fact that Theano is no longer updated, Keras has become the best tool for users to develop TensorFlow applications. Over the years, Keras and TensorFlow have also maintained a deep symbiotic relationship, further bringing more users into the TensorFlow ecosystem. "Keras Great God Returns: Comprehensive Evolution of Deep Learning! Implementing CNN, RNN, GRU, LSTM, GAN, VAE, Transformer in Python" was written by the founder of Keras himself, and no one knows the Keras set of tools better than him. However, this is not a Keras manual, but an introductory book that leads learners to explore deep learning from scratch, and then expand their understanding of deep learning. Learn about the standard operating procedure for solving machine learning problems, and how to overcome the problems you may encounter in practice in the future. Through the following selected content, let's first understand what interesting applications can be realized through Keras? How did the AI painting technology that went viral in the community develop? Recently, a group of developers created MidJourney, a brand new AI mapping tool. Compared with the current mainstream rendering tools, MidJourney is faster and of better quality, and it is completely impossible to tell whether it is generated by a computer or drawn by a human. The method of generating pictures is also very simple: as long as you input English keywords on MidJourney's server, AI will generate corresponding works, and the process only takes about 1 minute. Therefore, some people on the Internet jokingly said: "The technology that the painter has worked hard for 10 years can be done by AI in only 60 seconds." It is undeniable that the emergence of deep learning has extended many interesting technologies. In addition to the above image generation technologies, another type of technology that is more widely used is image modification technology . In the summer of 2015, Leon Gatys et al. published the technique of neural style transfer. The Neural Style algorithm has been improved many times since its publication, resulting in many functional changes, and has been introduced to many smartphone photo applications.
Website: https://techtuba.com/
Forums
Topics Started: 0
Replies Created: 0
Forum Role: Participant