Can a robot code my website?
summary
Algorithm replicates Tony Beltramelli’s pix2code paper and uses CNNs + LSTMs with softmax to map mockup images to DSL tokens, then to HTML/CSS. A Python/TensorFlow model is used for training and imported for prediction into a ML.NET implementation. Results produce correct buttons, links and layout tokens but consistently fails to place images because the training set lacks them. Outcome is workable for front-end scaffolding but quality hinges on the chosen training corpus.
AI-written summary · about
chapters (6)
- 0:00Introduction and context
- 1:09The design-to-code problem
- 2:57Deep learning basics
- 4:26Architecture of pix2code
- 7:00Implementation and demos
- 10:07Future potential and closing
related talks
29mAutomating Software Development with Deep Learning
46mCode + AI: Will Robots Take Our Coding Jobs? ML Applied to Programming
01h 04mFrom Pixels to Words: Exploring Multi-Modal AI with Python
28mHere Come The Robots - Django and Machine Learning
01h 55mCome As You Are and Learn to Code Your First AI
35mAI JavaScript Rocks
57mWhat if Picasso was a robot? Teaching art to the machine
34mMachine learning for front-end developers