Exogenous and Endogenous Data Augmentation for Low-Resource Complex Named Entity Recognition - Lecture 1
Explore data augmentation techniques for improving complex named entity recognition in low-resource scenarios, enhancing NLP model performance.
Building Transformer Tokenizers - Dhivehi NLP #1
Learn to build an effective WordPiece tokenizer for Dhivehi, a low-resource language, overcoming unique challenges in NLP application. Explore tokenizer components, implementation, and usage for this fascinating language.
CMU Multilingual NLP 2020 - Low Resource ASR
Explore low-resource ASR techniques, including lexicon-free systems, cross-lingual recognition, and speech translation. Learn innovative approaches for languages with limited data.
CMU Multilingual NLP - The LORELEI Project
Explore rapid development of low-resource information extraction in new languages, focusing on the LORELEI project's techniques, challenges, and real-world applications in disaster scenarios.
CMU Multilingual NLP - Information Extraction
Explore information extraction for low-resource languages and its application in disaster response through LORELEI project, with insights on methods, results, and practical implications.
CMU Multilingual NLP 2020 - Text to Speech
Explore text-to-speech systems for low-resource languages, covering analysis, pronunciation, prosody, waveform generation, and cross-lingual models. Learn efficient TTS building and multilingual challenges.
Low Resource Machine Translation
Explore machine translation techniques for low-resource languages, covering evaluation methods, supervised and unsupervised learning approaches, and domain adaptation. Gain insights into Nepali-English translation case study.
Generative AI for Social Good - Applications and Ethical Considerations
Explore generative AI's impact on social good, from medical imaging to language support. Learn about ethical applications in diversity training, bias detection, and enhancing communication in underrepresented languages.
Why GenAI Needs Careful Training Data Management
Explore the critical role of training data management in GenAI, including harmonization across stages and its impact on model effectiveness and safety in various domains and languages.
Low-resource Morphological Generation with Neural Sequence-to-Sequence Models
Explore neural sequence-to-sequence approaches for morphological generation in low-resource languages, including multi-task learning, cross-lingual transfer, and semi-supervised techniques.
Fine-tuning Llama 3 on Wikipedia Datasets for Low-Resource Languages
Learn to fine-tune Llama 3 on Wikipedia datasets for low-resource languages. Master dataset creation, LoRA setup, blending techniques, and parameter optimization for improved language model performance.
Extension is All You Need - But Sometimes Good Intentions Help Too
Explore innovative approaches to language extension and their impact on linguistic analysis, with insights on balancing intentions and outcomes.
Why You Should Study Your Native Language. | Ismail Abdirahman | TEDxCIU
When we lose our language, do we lose our soul? Here to champion the power of the native tongue... Your ability to express ideas clearly begins with your native language. Studying it strengthens articulation—how you form thoughts, choose words, and communicate meaning with precision. Ismail discus
The future of languages in the technological era: The Livonian case | Valts Ernštreits | TEDxTallinn
This talk explores the fragile yet powerful journey of the Livonian language, one of Europe’s most endangered tongues. Through stories of community, history, and resilience, it reveals how a language on the edge of disappearance can still be kept alive — and why preserving linguistic diversity matte