Search for tag: "shapes"

4 Media

Is it an Animal?: Reconstructing Ancient Fossils… duration 2 minutes 49 seconds Is it an Animal?: Reconstructing Ancient Fossils… Is it an Animal?: Reconstructing Ancient Fossils with Machine Learning, Advik Eswaran, UG '27 (7B7798F6)	Is it an Animal?: Reconstructing Ancient Fossils with Machine Learning, Advik Eswaran, UG '27 (7B7798F6) My research project was on generating a 3D reconstruction of ancient fossils (565 million years old) from Siberia. Working with over 2000 cross-sectional images of the fossil sample, which is encased… prd24 fossils animals sponges evidence shape sample rock change machine samples edges regularity fossil image process approach cross shapes climate explosion From Research Princeton Research Day 10 likes 84 plays Date May 3rd, 2024 Speaker Advik Eswaran, UG '27 (7B7798F6) Department Geosciences
Analyzing How Neural Networks Manipulate the… duration 2 minutes 47 seconds Analyzing How Neural Networks Manipulate the… Analyzing How Neural Networks Manipulate the Shape of Data, Reuel Williams, UG '24 (9C22A282)	Analyzing How Neural Networks Manipulate the Shape of Data, Reuel Williams, UG '24 (9C22A282) Understanding the mathematical shape of data that neural networks learn from can provide valuable insights into their inner workings. In my senior thesis, I delve into this concept using a… prd24 shape network digits images machine computer component task networks science donut coffee thing shapes dogs classification intelligence math areas dates From Research Princeton Research Day 3 likes 39 plays Date May 3rd, 2024 Speaker Reuel Williams, UG '24 (9C22A282) Department Mathematics
3D Object Reconstruction of Unseen and Unlabeled… duration 2 minutes 58 seconds 3D Object Reconstruction of Unseen and Unlabeled… 3D Object Reconstruction of Unseen and Unlabeled Point Clouds, Gene Chou, UG '22 (3963734)	3D Object Reconstruction of Unseen and Unlabeled Point Clouds, Gene Chou, UG '22 (3963734) Learning 3D representations of objects is a task at the heart of computer vision, robotic manipulation, scene understanding, medicine, and content generation. Implicit neural representations which… computer vision 3d objects 3d representation 3d reconstruction implicit neural representations signed distance functions (sdf) learn priors of 3d shapes episodic training object classes baseline points distance representations point training 3d detail surface sds method shapes stages baselines categories scale process scenarios From Research Princeton Research Day 0 likes 40 plays Date April 27th, 2022 Speaker Gene Chou Department Computer Science
Density Fluctuations and Hyperuniformity in Dense… duration 2 minutes 43 seconds Density Fluctuations and Hyperuniformity in Dense… Density Fluctuations and Hyperuniformity in Dense Packings of Superballs, Charles Maher, G4 (3959561)	Density Fluctuations and Hyperuniformity in Dense Packings of Superballs, Charles Maher, G4 (3959561) A packing is any collection of nonoverlapping objects. Dense, disordered packings are useful models of glasses, biological systems, and granular media. The study of dense packings of nonspherical… density fluctuations hyperuniformity superballs packing ordered disordered mechanically stable cube cubic octahedron octahedral symmetry granular thermal electrical conductivity spectral density orientation proportional density fluctuations materials values shapes slide objects bowl uniform length distance inverse conductivity properties scales property symmetries sphere particles From Research Princeton Research Day 1 likes 15 plays Date April 27th, 2022 Speaker Charles Maher Department Chemistry