PanoLlama: Generating Endless and Coherent Panoramas with Next-Token-Prediction LLMs

TL;DR

PanoLlama introduces an autoregressive framework for endless, coherent panoramic image generation, overcoming size limitations of previous models and enabling diverse applications with state-of-the-art coherence, fidelity, and aesthetics.

Contribution

It presents a novel training-free token redirection strategy for panoramic image generation with autoregressive models, allowing unlimited size and enhanced coherence.

Findings

Achieves state-of-the-art coherence (47.50%) in panoramic images.

Enables mask-free layout control and multi-guidance synthesis.

Supports endless panorama generation with improved fidelity and aesthetics.

Abstract

Panoramic Image Generation (PIG) aims to create coherent images of arbitrary lengths. Most existing methods fall in the joint diffusion paradigm, but their complex and heuristic crop connection designs often limit their ability to achieve multilevel coherence. By deconstructing this challenge into its core components, we find it naturally aligns with next-token prediction, leading us to adopt an autoregressive (AR) paradigm for PIG modeling. However, existing visual AR (VAR) models are limited to fixed-size generation, lacking the capability to produce panoramic images. In this paper, we propose PanoLlama, a novel framework that achieves endless and coherent panorama generation with the autoregressive paradigm. Our approach develops a training-free strategy that utilizes token redirection to overcome the size limitations of existing VAR models, enabling next-crop prediction in both horizontal and vertical directions. This refreshes the PIG pipeline while achieving SOTA performance in coherence (47.50%), fidelity(28.16%), and aesthetics (15%). Additionally, PanoLlama supports applications other PIG methods cannot achieve, including mask-free layout control, multi-scale and multi-guidance synthesis. To facilitate standardized evaluation, we also establish a dataset with 1,000 prompts spanning 100+ themes, providing a new testing benchmark for PIG research. The code is available at https://github.com/0606zt/PanoLlama.