PyPottery: an AI-powered end-to-end suite for pottery processing and publication

Problem

The paper addresses the labor-intensive post-production workflow for pottery documentation, which is often time-consuming and requires manual effort. The authors propose an AI-powered solution to streamline this process. As a preprint, it has not undergone peer review.

Method

The core technical contribution is the development of PyPottery, an end-to-end suite comprising several components:

• PyPotteryScan: This component automates image extraction and handwriting recognition, facilitating the digitization of pottery documentation.
• PyPotteryInk: It provides automatic inking of pencil drawings, enhancing the visual quality of the documentation.
• PyPotteryTrace: This module performs semantically-aware vectorization, converting raster images into vector formats while preserving semantic information.
• PyPotteryLayout: It generates automated layouts for the documentation, optimizing the presentation of the pottery drawings. The system was trained and evaluated using a dataset of 50 hand-drawn sheets containing 240 pottery drawings sourced from Terramara di Montale in Italy.

Results

The implementation of PyPottery yielded a median perceived speedup of 40× compared to traditional workflows, with a range of speedup reported between 17.5× and 120×. This significant improvement demonstrates the effectiveness of the proposed AI suite in enhancing the efficiency of pottery documentation processes.

Limitations

The authors do not report any limitations in the study. However, the absence of a peer review may imply that potential weaknesses or areas for improvement have not been critically evaluated.

Why it matters

The implications of this work are substantial for the field of pottery documentation and similar artistic domains. By automating labor-intensive tasks, PyPottery can reduce the time and effort required for documentation, allowing artists and researchers to focus on creative processes. Furthermore, the methodologies developed could be adapted for other forms of artistic documentation, potentially transforming workflows in various creative industries.