Multi-Agent Reinforcement Learning for Indoor Monitoring
Enhance dynamic indoor environment monitoring with cooperative sensing using multi-agent reinforcement learning for improved accuracy and adaptability.
Optimizing Multi-Node MoE Inference with Expert Activation
Discover strategies to improve multi-node Mixture-of-Experts inference by balancing expert load and reducing communication overhead for faster LLM performa...
UpstreamQA: Modular Framework for Video Question Answering
Discover UpstreamQA, a modular framework enhancing explicit reasoning and interpretability in video question answering tasks for better AI performance.
UNSEEN: Defense Against AR-LLM Social Engineering Attacks
Discover UNSEEN, a cross-stack defense framework protecting against AR-LLM social engineering attacks by enhancing access control and LLM unlearning.
Hybrid CNN-ViT Model with Adaptive Attention for Brain Tumor MRI
Discover a hybrid CNN-ViT deep learning model with adaptive attention gate that boosts brain tumor MRI classification accuracy to 97.6%.