https://openreview.net/forum?id=FKi6yjXwCN&referrer=%5Bthe%20profile%20of%20Ang%20Cao%5D(%2Fprofile%3Fid%3D~Ang_Cao1)
LOCATE 3D: Real-World Object Localization via Self-Supervised Learning in 3D | OpenReview
We present LOCATE 3D, a model for localizing objects in 3D scenes from referring expressions like "the small coffee table between the sofa and the lamp."...
real world objectself supervised learning
https://www.amazon.science/publications/bridging-the-gap-to-real-world-object-centric-learning
Bridging the gap to real-world object-centric learning - Amazon Science
Humans naturally decompose their environment into entities at the appropriate level of abstraction to act in the world. Allowing machine learning algorithms to...
bridging the gapreal world objectlearning amazon
https://openreview.net/forum?id=b9tUk-f_aG
Bridging the Gap to Real-World Object-Centric Learning | OpenReview
Our method uses slot attention with self-supervised DINO features to discover objects on real-world data.
bridging the gapreal world objectcentriclearningopenreview
https://deepai.org/publication/bridging-the-gap-to-real-world-object-centric-learning
Bridging the Gap to Real-World Object-Centric Learning | DeepAI
Sep 29, 2022 - 09/29/22 - Humans naturally decompose their environment into entities at the appropriate level of abstraction to act in the world. Allowing m...
bridging the gapreal world objectcentriclearningdeepai
https://openreview.net/forum?id=zRYSJbcRcV
Stanford-ORB: A Real-World 3D Object Inverse Rendering Benchmark | OpenReview
We introduce Stanford-ORB, a new real-world 3D Object inverse Rendering Benchmark. Recent advances in inverse rendering have enabled a wide range of real-world...
a real3d objectinverse renderingstanfordorb