Examining the performance of automatic speech recognition for deaf and hard of hearing users.
― 11 min read
Computer Science - Computation and Language
RSSA new method helps identify test data contamination in LLMs using token probabilities.
― 8 min read
Examining the growth and importance of automated systems for fact-checking claims.
― 6 min read
New framework uses AI to enhance error detection in 3D printing.
― 6 min read
Research explores AI's role in enhancing epidemic monitoring using diverse data sources.
― 5 min read
A new model transforms plain texts into fitting song lyrics.
― 6 min read
This study examines the interaction of knowledge from LLMs and semantic towers.
― 6 min read
This study analyzes how diphthongs and monophthongs differ in production and movement.
― 5 min read
This study examines how users' mental models affect their interactions with task-oriented dialog systems.
― 6 min read
Explore how dual encoders connect images to text.
― 8 min read
A novel method accounts for human disagreement in natural language processing tasks.
― 7 min read
ARWFML allows easy creation of AR applications without deep technical knowledge.
― 6 min read
New method enhances ASR accuracy using language models for better transcriptions.
― 4 min read
AI chatbots improve accuracy in diagnosing depression through advanced data analysis.
― 6 min read
SurGen creates high-quality surgical videos for better training.
― 5 min read
A new system corrects speaker identification errors for clearer conversation transcripts.
― 7 min read
Study shows LLMs struggle with multiple-choice questions lacking correct options.
― 4 min read
This article examines the progress of vision-language models and their reasoning capabilities.
― 4 min read
A deep dive into how next-token prediction shapes language understanding in models.
― 6 min read
X-Codec improves audio generation by integrating semantic understanding into processing.
― 6 min read
A look at the strengths and challenges of using COMET for machine translation assessment.
― 6 min read
A new approach streamlines safety and helpfulness in language model training.
― 9 min read
This study introduces a model to identify populist rhetoric in political discourse.
― 8 min read
Combining AI and human annotations improves data accuracy and efficiency in research.
― 6 min read
This article examines how to measure factual accuracy in language model summaries.
― 6 min read
A new method focuses on relationships to enhance language model accuracy.
― 5 min read
BaichuanSEED shares its data processing methods for better language model development.
― 4 min read
A look at LLM development and challenges for EU languages.
― 6 min read
A new framework improves answer accuracy in AI models by focusing on evidence.
― 5 min read
MRSE enhances online shopping by integrating text and image data for better search results.
― 5 min read
This article discusses a new method for generating synthetic data using various models.
― 7 min read
A system to organize insights from social media posts effectively.
― 6 min read
Introducing a method that enhances model understanding of long texts.
― 6 min read
Introducing VHAKG, a tool connecting synchronized videos and knowledge for research.
― 6 min read
A new method streamlines aligning language models with human preferences.
― 5 min read
ToxDet proposes a new method to identify harmful outputs in language models.
― 5 min read
A new benchmark assesses language model performance in AAVE vs. Standard American English.
― 6 min read
A study on how different descriptions impact machine learning models in medical imaging.
― 5 min read
This method enhances the learning process of knowledge graphs through structured training.
― 5 min read
A new tool helps users understand complex privacy policies easily.
― 6 min read