PixelBytes: Catching Unified Embedding for Multimodal Generation

TL;DR

PixelBytes introduces a unified embedding approach for multimodal data, enabling coherent sequence generation across text and images by leveraging various model architectures and innovative embedding techniques.

Contribution

The paper presents PixelBytes Embedding, a novel unified representation method for multimodal data, integrating diverse inputs into a single cohesive embedding for sequence generation.

Findings

Bidirectional models with PxBy embedding generate coherent multimodal sequences.

PixelBytes achieves integrated understanding and generation of text and pixelated images.

Experimental results on the PixelBytes Pokémon dataset validate the approach.

Abstract

This report introduces PixelBytes Embedding, a novel approach for unified multimodal representation learning. Our method captures diverse inputs in a single, cohesive representation, enabling emergent properties for multimodal sequence generation, particularly for text and pixelated images. Inspired by state-of-the-art sequence models such as Image Transformers, PixelCNN, and Mamba-Bytes, PixelBytes aims to address the challenges of integrating different data types. We explore various model architectures, including Recurrent Neural Networks (RNNs), State Space Models (SSMs), and Attention-based models, focusing on bidirectional processing and our innovative PxBy embedding technique. Our experiments, conducted on a specialized PixelBytes Pok{\'e}mon dataset, demonstrate that bidirectional sequence models with PxBy embedding and convolutional layers can generate coherent multimodal sequences. This work contributes to the advancement of integrated AI models capable of understanding and generating multimodal data in a unified manner.