Raft
PulseAugur coverage of Raft — every cluster mentioning Raft across labs, papers, and developer communities, ranked by signal.
4 day(s) with sentiment data
-
New method rectifies disparity for 360 stereo images
Researchers have developed a method to improve disparity estimation for 360-degree stereo images by using a spherical-to-equirectangular (ERP) projection. This preprocessing step straightens epipolar curves, allowing fo…
-
RAG vs. Fine-Tuning: Choosing the Right LLM Approach for Knowledge vs. Behavior
The debate between Retrieval-Augmented Generation (RAG) and fine-tuning for LLMs hinges on whether the goal is to impart new knowledge or alter the model's behavior. RAG is presented as the superior method for injecting…
-
New RAFT framework refines domain fine-tuning, reduces model forgetting
Researchers have introduced RAFT, a novel two-stage framework designed to improve domain-specific fine-tuning of language models while mitigating performance degradation on general tasks. RAFT addresses issues like supe…
-
SMART method boosts soccer player pose estimation by 38.6%
Researchers have developed a new method called SMART for estimating the 3D poses of soccer players from broadcast video. This approach fine-tunes the SMPLest-X model and incorporates RAFT dense optical flow tracking, al…
-
New method tackles dynamic object segmentation in turbulence
Researchers have developed a novel approach for dynamic object segmentation in turbulent conditions, specifically for the CVPR 2026 UG2+ Challenge Track 3. Their method, which requires no end-to-end training, integrates…
-
New algorithm boosts optical flow estimation efficiency
Researchers have developed a novel algorithm to improve the efficiency of optical flow estimation, a computer vision task crucial for understanding motion in images. This new method significantly reduces the computation…
-
New Hybrid AI Architecture Enhances Wind Turbine Blade Inspection
Researchers have developed a novel hybrid architecture for automated industrial inspection, specifically for wind turbine blade maintenance. This system integrates a vision model for defect localization with a language …
-
Hybrid AI pipeline automates wind turbine blade defect reporting
Researchers have developed a novel pipeline for automated industrial inspection, specifically for wind turbine blades. This system integrates a vision model for defect localization with a language model for generating s…
-
New optical flow method skips test-time scaling using foundation models
Researchers have developed a new method for estimating dense optical flow that bypasses the need for computationally intensive test-time scaling. This approach leverages pretrained foundation models, specifically DINO-v…
-
Data Scientist explores machine unlearning measurement and achievement
This article explores the concept of machine unlearning, focusing on methods to measure and achieve the "forgetting" of specific data within AI models. The author, a Data Scientist at Raft, draws upon a conference prese…