We are excited to get back to work, to revive Data Phoenix from the ashes of war. Wait a bit, and we’ll resume the stream of interesting news & updates on all things AI/ML and Data.
As to the war, it continues. From now on, our team will spend 50% of our ad revenue to support the armed forces of Ukraine — to, finally, end suffering of the Ukrainian people. You too can add your voice to support Ukraine by promoting your brand with us. Learn more about the opportunities here.
If you want to post your articles, research, or any other type of content on our website or in the digest, you can always reach out to us to discuss the terms. Don’t hesitate!
Deploying Models from Dev to Production with MLflow and Bridge
In this article, you’ll find a step-by-step guide on using MLflow and Bridge to deploy ML models and implement the machine learning workflow, from development to production.
Emerging Architectures for Modern Data Infrastructure
This comprehensive guide includes the input from dozens of data experts. If you want to learn the latest from the data infrastructure landscape, make sure you check it out!
How to Structure a Data Science Project for Readability and Transparency
In this article, you will find an easy-to-use template that incorporates best practices of structuring data science projects, to help you organize your data science workflow.
Backpropagation in RNN Explained
In this article, you’ll find a step-by-step explanation of computational graphs and backpropagation in a recurrent neural network. A simple and clear 101 for beginners.
Correcting Text Orientation with Tesseract and Python
Image preprocessing is an essential component of any OCR system. The most important pre-processing step is text orientation. Learn how to handle it more efficiently.
Vision Transformers for Single Image Dehazing
DehazeFormer is a new vision transformer for single image dehazing that consists of the modified normalization layer, activation function, and spatial information aggregation scheme.
Unsupervised Image-to-Image Translation with Generative Prior
In this paper, you’ll learn about a novel framework that leverages the generative prior from pre-trained class-conditional GANs to learn rich content correspondences across various domains.
Colossal-AI: A Unified Deep Learning System For Large-Scale Parallel Training
Colossal-AI is a unified parallel training system designed to seamlessly integrate different paradigms of parallelization techniques to help the AI community write distributed models.
MaskGIT: Masked Generative Image Transformer
This paper proposes a novel image synthesis paradigm using a bidirectional transformer decoder — MaskGIT. It outperforms the state-of-the-art transformer model on the ImageNet dataset.
Pre-Trained Language Models for Interactive Decision-Making
The researchers describe a framework for imitation learning that enables effective combinatorial generalization across different environments, such as VirtualHome and BabyAI.
CLIPasso: Semantically-Aware Object Sketching
CLIPasso is a method for performing CLIP-guided Semantically-Aware Object Sketching that converts an image of an object to a sketch while preserving its key visual features.
Block-NeRF: Scalable Large Scene Neural View Synthesis
Block-NeRF is a variant of Neural Radiance Fields that represent large-scale environments. It is the largest neural scene representation capable of rendering an entire neighborhood of San Francisco.
FILM: Frame Interpolation for Large Motion
FILM is a frame interpolation algorithm that synthesizes multiple intermediate frames from two input images with large in-between motion. It outperforms state-of-the-art methods on the Xiph large motion benchmark.
AlphaCode Explained: AI Code Generation
AlphaCode is DeepMind's new language model for generating code. The field of NLP within AI and ML has exploded so this video can help you understand how it works for code generation.
Data Phoenix Newsletter
Join the newsletter to receive the latest updates in your inbox.