About this Abstract |
Meeting |
2025 Annual International Solid Freeform Fabrication Symposium (SFF Symp 2025)
|
Symposium
|
2025 Annual International Solid Freeform Fabrication Symposium (SFF Symp 2025)
|
Presentation Title |
GenAI-AM-Bench: A Multimodal Benchmark Dataset for Evaluating Generative AI Tools in Metal Additive Manufacturing |
Author(s) |
Nowrin Akter Surovi, Yeun Park, Paul Witherell |
On-Site Speaker (Planned) |
Nowrin Akter Surovi |
Abstract Scope |
Generative artificial intelligence (GenAI) models have the potential to advance and automate processes in additive manufacturing (AM). However, the effectiveness of existing GenAI models in the AM domain is not yet well understood. It is important to create a dedicated dataset that can be used to evaluate and benchmark GenAI tools for AM tasks. In this paper, we propose GenAI-AM-Bench, a multi-modal benchmark data set designed to evaluate the performance of GenAI tools in metal additive manufacturing tasks. The dataset includes both textual and image data. The textual content is organized into four categories: design, pre-process, in-situ monitoring, and post-process. It also includes different types of questions, like free-form, multiple-choice (MCQs), and true/false, to understand the capabilities of GenAI tools in AM-specific contexts. By organizing the information clearly, GenAI-AM-Bench provides a valuable resource to benchmark existing GenAI tools and help improve research in using AI for metal additive manufacturing. |
Proceedings Inclusion? |
Planned: Post-meeting proceedings |