Discover the journey of developing a Deep Learning-enhanced Android app in this 57-minute conference talk from ML Conference 2017. Learn from Alexander Frank and Andreas Eberle of arconsis IT-Solutions GmbH as they share valuable insights and experiences gained during their project. Explore the decision-making process between deep learning and traditional approaches, best practices for data collection and labeling, and the implementation of efficient production-ready apps using Google TensorFlow API. Gain practical knowledge on AI neural networks, image recognition techniques, and overcoming development constraints. Follow their step-by-step process from project inception to deployment, including data preprocessing, classical computer vision methods, and machine learning integration. Benefit from their lessons learned to avoid common pitfalls and streamline your own deep learning app development journey.
Overview
Syllabus
Intro
Company Introduction
Project Description
Project Experience
Where to begin
AI Neural Networks
Collecting Data
Making Sense of Data
Training
Process
Cascade
Preprocessing
Classical CV
Different color space
Machine learning to neural networks
Image recognition
Postprocessing
Constraints
Questions
Taught by
MLCon | Machine Learning Conference
