Events¶
CLAMS project events, presentations, and conference appearances.
FantasticFutures 2026¶
2026-09-15 - We presented three pieces of work with GBH Archives at Fantastic Futures 2026, hosted by the Library of Congress in Washington, DC:
“Building AI Tools with Communities”: how Samoan and Yup’ik language fellows shape the data, tooling, and workflow behind spoken language identification for public media archives.
“Smart Segmentation to Support More Transparent Audiovisual Description”: three open-source CLAMS apps that split long programs at natural content breaks so that LLM summaries stay small, checkable, and navigable.
“On-demand Agents vs. Bulk Processing for RAG”: a benchmark of over 1,000 questions on archival video comparing precomputed retrieval against agents that decide at query time which CLAMS tools to run.
The language identification work received the 2026 AI4LAM Innovation Award for Best Research Project. Recordings of the talks will be linked here once they are published.
IASA Journal 2026¶
2026-09-12 - Our IASA 2025 talk on Hawaiian chyrons is now a journal article. Understanding On-Screen Text: Do AI Tools Struggle with Hawaiian Chyrons? appears in IASA Journal no. 56 and shows how Hawaiian names, orthography, and visual elements disproportionately trip up OCR and vision-language models.
AmericasNLP 2026¶
2026-07-03 - Our pilot study on spoken language identification for low-resource indigenous languages, surveying how existing systems cover Central Alaskan Yup’ik and Samoan in US public broadcast archives, was published at the AmericasNLP workshop at ACL 2026.
SIGUL 2026¶
2026-05-11 - Our paper on structured entity extraction from Hawaiian television chyrons with vision-language models was published at the SIGUL workshop at LREC 2026.
FantasticFutures 2025¶
2025-12-04 - Together with GBH Archives, we presented “Surfacing and Tracing Contributors in Large Video Collections” at the Fantastic Futures 2025 conference, showing how CLAMS-based chyron extraction feeds a human-in-the-loop cataloging workflow that traces on-screen contributors across decades of public television.
IASA 2025¶
2025-09-10 - We had two presentations at the IASA 2025 conference: one on spoken language identification for low-resource languages and another on on-screen text understanding.
C&C 2025¶
2025-06-22 - Our paper A Platform for AI-assisted Archival Metadata Generation was presented at C&C 2025, part of HCI International 2025. It gives an up-to-date account of the CLAMS platform and the MMIF interchange format, and walks through real cataloging workflows and evaluations built with the American Archive of Public Broadcasting.
code4lib 2025¶
2025-03-10 - A presentation on a CLAMS app for image classification at code4lib 2025.
MMM 2025¶
2025-01-08 - We are presenting our paper on multimodal interoperability at the 2025 International Conference on Multimedia Modeling (MMM).
CAS 2024¶
2024-12-17 - Our work on LLM-based video summarization is being presented at the Computational Archival Science (CAS) Workshop at IEEE Big Data 2024.
AMIA 2024¶
2024-12-06 - We gave a presentation at the AMIA 2024 conference on the combination of LLM and ASR for video understanding.
LREC 2022¶
2022-06-20 - Our paper on designing data formats for CLAMS app interoperability was presented at the 2022 Language Resources and Evaluation Conference (LREC).
IIIF 2021¶
2021-06-23 - We presented CLAMS and how we are using IIIF for visualizing MMIF at the IIIF 2021 conference.
AMIA 2020¶
2020-11-12 - CLAMS was featured in a workshop at the AMIA 2020 conference.
FantasticFutures 2019¶
2019-12-05 - We presented CLAMS at a workshop at the Fantastic Futures 2019 conference. In addition to the workshop, we presented a lightning talk about CLAMS and the American Archive of Public Broadcasting.
JTS 2019¶
2019-10-05 - A workshop on CLAMS was held at the JTS 2019 conference.
ACL-SIGHUM LaTeCH-CLfL 2019¶
2019-06-07 - Our paper was accepted at the LaTeCH-CLfL workshop at NAACL 2019.