All Articles

Beyond Unimodal: Embracing Multimodal Deep Learning for Enhanced AI

Multimodal deep learning is a subfield of machine learning that combines multiple sources of data to gain understanding.

Understanding CLIP

The reason why CLIP is so effective in zero-shot learning is because of the contrastive learning approach to jointly train a neural network.

Leveraging Reasoning-Modulated Representations for MRI and CT scan: A case study

Reasoning-modulated representations aim to create AI systems that are not just data-driven, but also capable of understanding and reasoning.

NAS + Reinforcement learning: A review

Neural architecture search (NAS) is a method for automatically discovering the best neural network architecture for a given task.

Perplexity in AI and Humans

Perplexity is a useful metric to evaluate models in Natural Language Processing (NLP).

Attention Probabilistic U-net

A new hybrid method to find anomalies in CT scans (on-going experiment) Image classification is not new to us and the same stands true...

Artificial intelligence for designing new materials

Understanding the material helps us to understand the working of the universe.

Cost-efficient method to predict 3D hemodynamics

The deep learning methods are inexpensive compare to modeling a cardiovascular hemodynamic system.

OpenAI's GPT-3 Language Model

GPT-3 is now trending as one of the most powerful language models in the world.

AlphaFold

On 8 November 2020, Deepmind of Google an AI created that could plausibly understand the biological machines — Proteins.

Generative Methods in Medical Applications

Generative adversarial networks (GANs) have achieved state-of-the-art results in recent years and it is able to produce realistic images .

A Note on Bayesian Optimisation

Sometime during the 1740s, Reverend Thomas Bayes made an ingenious statistical discovery, an idea, a formula that had changed the world

1/2

PerceptronAI