New Framework for Nighttime Drone Localization: Thermal Imaging + Semantic 3D Maps
A latest arXiv paper proposes the "Lights Out" framework, which aligns thermal infrared images with semantic 3D maps to …
📂 Research
Latest articles in Research
Google Releases TurboQuant: A New Efficient KV Cache Compression Solution
Google has launched the TurboQuant algorithm suite and open-source library, focused on advanced quantization and compres…
Breakthrough in Three-Vehicle Platooning Platform: Tackling Path Tracking Challenges in Cooperative Driving
A latest research paper proposes a scaled three-vehicle platooning experimental platform that addresses the lateral devi…
FalconApp: Turning Your iPhone Into a Robot Perception Terminal in Seconds
A research team has launched FalconApp, which requires only a brief iPhone scan to automatically generate object detecti…
Diffusion Language Models Enter the Speech Recognition Arena
New research explores applying Masked Diffusion Language Models (MDLM) and Uniform State Diffusion Models (USDM) to auto…
Breakthrough in Uncertainty-Aware Offline Data-Driven Multi-Objective Optimization
A latest arXiv study proposes an uncertainty-aware offline data-driven multi-objective optimization method that improves…
Study Reveals LLMs 'Playing Dumb' Is Positional Collapse, Not Answer Avoidance
A new preregistered study using option-order randomization experiments found that when large language models are prompte…
PBiLoss: Eliminating Popularity Bias in Recommendation Systems Through Regularization
Researchers propose PBiLoss, a popularity-aware regularization technique that directly mitigates popularity bias during …
Neural Assemblies Can Learn Causal Direction, Marking a Key Step Forward for Brain-Inspired Computing
A new arXiv study shows that Neural Assemblies can autonomously learn the causal directionality between variables throug…
Plain Text Data Boosts Performance of Encoder-Dominant Speech Recognition Models
New research explores how to efficiently leverage plain text data to improve encoder-dominant speech recognition models.…
Evolutionary Algorithms Drive New Breakthrough in Spiking Neural Network Feature Selection
A new study applies the biologically realistic JASTAP spiking neural network model to classification tasks, combining ev…
ARIEL Robot System: Spatial Mating Mechanisms Drive a New Paradigm in Embodied Evolution
A research team proposes a spatially embedded evolutionary algorithm that enables gecko-inspired quadruped robots to aut…