Abstract Concept Recognition for Advanced Video Understanding
Explore how AI recognizes abstract concepts in videos, enhancing understanding beyond objects with foundation models and semantic analysis.
Once4All: Efficient SMT Solver Fuzzing with LLM Generators
Discover Once4All, a framework using LLM-synthesized generators for efficient, accurate SMT solver fuzzing, improving bug detection and testing speed.
ShadowNPU: Efficient NPU-Based On-Device LLM Inference
Discover ShadowNPU's system and algorithm co-design for efficient, privacy-preserving on-device LLM inference using NPU-centric techniques.
Optimizing AutoML Pipelines with In-Context Decision Making
Discover how in-context decision making enhances AutoML pipeline optimization using PS-PFN, improving model selection and adaptation efficiency.
Estimating Target Task Performance in Semi/Self-Supervised Learning
Discover a low-cost method to estimate target task performance from unsupervised pretext tasks in semi/self-supervised learning for improved AI models.