Generalized Decoding for Pixel, Image, and Language

TL;DR

X-Decoder is a unified model that seamlessly integrates pixel-level segmentation and language understanding, enabling versatile vision-language tasks with strong transferability and state-of-the-art performance across multiple datasets.

Contribution

It introduces a novel generalized decoding framework supporting all image segmentation types and vision-language tasks in a single model without pseudo-labeling.

Findings

Achieves state-of-the-art results on open-vocabulary segmentation and referring segmentation.

Demonstrates strong transferability to various downstream tasks in zero-shot and finetuning settings.

Offers flexible and efficient finetuning for diverse vision-language applications.

Abstract

We present X-Decoder, a generalized decoding model that can predict pixel-level segmentation and language tokens seamlessly. X-Decodert takes as input two types of queries: (i) generic non-semantic queries and (ii) semantic queries induced from text inputs, to decode different pixel-level and token-level outputs in the same semantic space. With such a novel design, X-Decoder is the first work that provides a unified way to support all types of image segmentation and a variety of vision-language (VL) tasks. Further, our design enables seamless interactions across tasks at different granularities and brings mutual benefits by learning a common and rich pixel-level visual-semantic understanding space, without any pseudo-labeling. After pretraining on a mixed set of a limited amount of segmentation data and millions of image-text pairs, X-Decoder exhibits strong transferability to a wide range of downstream tasks in both zero-shot and finetuning settings. Notably, it achieves (1) state-of-the-art results on open-vocabulary segmentation and referring segmentation on eight datasets; (2) better or competitive finetuned performance to other generalist and specialist models on segmentation and VL tasks; and (3) flexibility for efficient finetuning and novel task composition (e.g., referring captioning and image editing). Code, demo, video, and visualization are available at https://x-decoder-vl.github.io.