CVCLLGJun 15, 2022

A Unified Sequence Interface for Vision Tasks

arXiv:2206.07669v2187 citationsh-index: 162
Originality Highly original
AI Analysis

This work addresses the proliferation of distinct architectures in computer vision, offering a unified approach that could simplify model development and deployment.

The authors tackled the lack of a unified modeling framework in computer vision by proposing a pixel-to-sequence interface that formulates diverse tasks like object detection and image captioning as sequences of discrete tokens, achieving competitive performance compared to task-specific models.

While language tasks are naturally expressed in a single, unified, modeling framework, i.e., generating sequences of tokens, this has not been the case in computer vision. As a result, there is a proliferation of distinct architectures and loss functions for different vision tasks. In this work we show that a diverse set of "core" computer vision tasks can also be unified if formulated in terms of a shared pixel-to-sequence interface. We focus on four tasks, namely, object detection, instance segmentation, keypoint detection, and image captioning, all with diverse types of outputs, e.g., bounding boxes or dense masks. Despite that, by formulating the output of each task as a sequence of discrete tokens with a unified interface, we show that one can train a neural network with a single model architecture and loss function on all these tasks, with no task-specific customization. To solve a specific task, we use a short prompt as task description, and the sequence output adapts to the prompt so it can produce task-specific output. We show that such a model can achieve competitive performance compared to well-established task-specific models.

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes