We are pleased to announce the Fifth Workshop on Multimodal Machine Learning in Low-Resource Languages (MMLow @ AACL-IJCNLP 2026), held in conjunction with the 5th Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics and the 15th International Joint Conference on Natural Language Processing. MMLow 2026 will be conducted entirely online, with all workshop sessions and presentations taking place virtually. No in-person attendance is required. This cutting-edge event aims to explore the intersection of multimodal learning and low-resource languages. Multimodal learning focuses on integrating text, speech, images, and video to capture complex linguistic and social patterns. It has enabled significant progress across tasks such as content understanding, generation, and reasoning by merging complementary signals across different modalities. Recent advances in multimodal foundation models, large language models (LLMs), and large vision models (LVMs) have further accelerated this progress. However, low-resource and underrepresented languages, including indigenous, tribal, and endangered languages, remain critically underserved due to data scarcity, limited multimodal benchmarks, and inadequate evaluation frameworks.

MMLow 2026 aims to address these challenges by bringing together researchers and practitioners at the intersection of multimodal machine learning, NLP, and inclusive AI. The workshop emphasizes language diversity, content safety, and societal impact and provides a focused forum to discuss novel methods, datasets, evaluation strategies, and real-world deployments of multimodal AI systems. Building on the success of previous editions, MMLow 2025 @ SPELLL 2025, MMLow 2024 @ SPELLL 2024, MMLow 2023 @ SPELLL 2023, and MMLow 2022 @ ICON 2022, this workshop continues our exploration of multimodal machine learning and its applications to low-resource languages. By examining key challenges and advancing multimodal learning techniques, the workshop aims to establish a collaborative forum for scholars and researchers committed to strengthening low-resource language research. 

Themes and Topics of Interest:

• Multimodal LLM Agents for Low-Resource Languages

• Safe and Responsible Agent-Based Multimodal AI

• Generative and Agentic Multimodal AI

• Deployable Multimodal Agents for Real-World and Domain-Specific Applications

• Addressing data scarcity in low-resource languages

• Leveraging multimodal data fusion and integration for social media analytics

• AI-driven solutions for linguistic diversity

• Online content moderation in low-resource languages on social media

• Computational linguistics perspectives for low-resource and multimodal settings

• Social media data analytics for crisis management

• Performance evaluation of NLP and computer vision tasks in low-resource languages

• Multimodal benchmark creation and evaluation frameworks for underrepresented languages

• Ethical, inclusive, and trustworthy multimodal AI for low-resource communities

• Speech, vision, and text alignment for multilingual and low-resource applications

• Multimodal misinformation detection and harmful content analysis in low-resource languages

• Human-centered multimodal AI systems for societal impact and accessibility

Publication

MMLow 2026 will be held entirely virtually as part of AACL-IJCNLP 2026. All workshop presentations and sessions will take place online.

Code of Conduct

Our workshop follows the ACL Anti-Harassment Policy. We are committed to providing a respectful and harassment-free environment for all participants. Anyone who experiences harassment or hostile behaviour may contact a member of the ACL Professional Conduct Committee. Concerns will be treated confidentially.

Reference:

• https://www.aclweb.org/adminwiki/index.php?title=Anti-Harassment_Policy

• https://2026.aaclnet.org/participants/anti_harassment/