Search papers, labs, and topics across Lattice.
University of Zaragoza
7
0
7
Achieving RGB-D performance with only monocular input, MyGO-Splat revolutionizes SLAM by integrating closed-loop geometric feedback for enhanced scale stability.
HarmVideoBench reveals that existing benchmarks miss critical layers of harmful video understanding, while a new method boosts model accuracy by over 20%.
State-of-the-art image-to-video models are alarmingly vulnerable to visual prompt attacks, achieving up to 100% success in triggering harmful outputs.
Alpha-blending, a core optimization in 3D Gaussian Splatting, subtly hobbles feature learning, but a geometry-weighted fusion approach can unlock more accurate and efficient visual localization.
Forget training monolithic policies: RoboRouter intelligently routes robotic manipulation tasks to the best off-the-shelf policy, boosting success rates by 13% in the real world.
By "dreaming" plausible scene completions, Dream-SLAM enables robots to navigate dynamic environments more effectively, achieving better localization, mapping, and exploration than existing methods.
Certifiably optimal solutions to 3D vision problems are now within reach, but choosing the right global solver (BnB, CR, or GNC) requires navigating a complex trade-off between optimality, robustness, and scalability.
