Search papers, labs, and topics across Lattice.
8
0
12
Occlusion-aware attention can dramatically enhance panoptic segmentation performance without incurring significant computational costs.
Super-resolution techniques can significantly enhance image quality but may compromise defect detection, with advanced models performing worse than simpler methods in real-world applications.
GAIA redefines online data selection for LLM instruction tuning, achieving superior performance by dynamically prioritizing high-utility samples across the entire semantic space.
By decoupling trend and residual information, PMDformer achieves unprecedented accuracy in long-term time series forecasting, setting a new benchmark in the field.
Language models face irreducible error floors that prevent them from being universal solvers, revealing the critical limits of prompt-based learning.
Achieving an overall score of 84.76, DreamX-World 1.0 sets a new benchmark for interactive video generation, outperforming leading models in both camera control and visual fidelity.
Grounded action-centric evidence can dramatically enhance sentiment prediction in video ads by linking emotional cues to explicit visual events.
LLM agents struggle in dynamic environments, but EvoMem boosts their performance by capturing the evolution of memory, leading to better adaptability.
