Entity Type
Publication (70982)
Thesis (35393)
Project (5437)
Researcher (3155)
Multimedia (1794)
Open Course (448)
Issue (263)
Biography (112)
Volume (110)
Patent (89)
Album (11)
Category (9)
Research Data (7)
Proceedings Book (6)
Journal (3)
Conference (2)

Has File(s)
No (69588)
Yes (48234)

Author
Zhang, J. (1154)
Wang, J. (1141)
Bocci, A. (1124)
Kim, H. (1122)
Xie, S. (1108)

Subject
Engineering and Technology (3323)
Social Sciences and Humanities (2109)
Condensed Matter Physics (1376)
Natural Sciences (1145)
Nuclear and High Energy Physics (919)

Date Issued
1950 - 1959 (34)
1960 - 1979 (3598)
1980 - 1999 (14036)
2000 - 2019 (73018)
2020 - 2027 (25737)

Item Type
Journal Article (33823)
Master Thesis (28653)
Conference Paper (21712)
Ph.D. Thesis (6722)
Project (4264)

Realtime Access Map

Recently Added

A medieval city in transformation: A GIS-based documentation, identification, and assessment of the Byzantine heritage in Trebizond (Trabzon)
Sur, Selin; Serin, Ufuk; Conservation of Cultural Heritage in Department of Architecture (2026-7-30)
Trebizond (modern Trabzon) is one of the longest continuously inhabited cities on the Black Sea coast of Türkiye, with origins dating back to at least the Hellenistic period (330–30 BCE). It is also among the rare urban centers that seamlessly transitioned through the ‘Dark Ages’ (7th–9th centuries) into modern times without major settlement interruption. After the capture of Constantinople (Istanbul) by the Fourth Crusade in 1204, the city became the capital of one of the three Byzantine splinter states, n...
LEARNING HIERARCHICAL TASK-SPACE CONTROL POLICIES FOR WHOLE-BODY LOCOMOTION
Banuş, Ömer Tarık; Ankaralı, Mustafa Mert; Ali Emre , Turgut; Department of Robotics (2026-7-22)
This thesis presents a hierarchical quadruped locomotion controller in which a reinforcement-learning policy sits above a structured optimization-based whole-body controller (WBC). The policy generates control references for ground reaction forces, body movement, gait and footstep planning behaviors. Hence, the policy is able to affect all high level locomotion tasks to improve standard behavior. Whereas, by preserving the low-level WBC, the policy does not need to learn complex kinematics, dynamics, and co...
Generative super-resolution of 3D Gaussian splat sets
Bal, Batuhan; Cinbiş, Ramazan Gökberk; Department of Computer Engineering (2026-9-03)
3D Gaussian Splatting (3DGS) enables real-time novel view synthesis, but Gaussian representations reconstructed from low-resolution observations are typically sparse and lack fine geometric and appearance details, limiting the quality of high resolution novel-view rendering. Existing 3DGS super-resolution methods either refine each scene through test-time optimization with external 2D priors, or learn feed-forward mappings from low-resolution images to high-resolution splats. Neither line of work learns a r...
Topological Full Groups: Finite-Stage Approximations, Free Subgroups, and Substitution Models
Özdemir, Anıl; Benli, Mustafa Gökhan; Department of Mathematics (2026-8-24)
Topological full groups describe orbit structure in Cantor dynamics through locally defined transformations. For every infinite minimal subshift, we construct finitely presented finite-stage groups converging to its commutator subgroup as marked groups. We prove that these groups, and every finitely presented group surjecting onto the commutator subgroup, contain a non-abelian free subgroup. Hence, the amenable commutator subgroup is not finitely presented. We also study the Lysenok substitution subshift, d...
Acceleration of finite rate chemistry in laminar reacting flow simulations using an artificial neural network
Göçer, Furkan; Karaca, Mehmet; Department of Aerospace Engineering (2026-8-14)
This thesis develops an artificial neural network-based chemistry surrogate for accelerating chemical source term computation in reacting flow computational fluid dynamics. The proposed method replaces direct integration of stiff chemical kinetics with neural networks trained to predict reactive species mass fraction changes over a fixed time step. The approach is evaluated using a compact H2/air mechanism and a more complex Jet-A/air mechanism. The trained networks are integrated into a finite volume react...
View More

Communities in OpenMETU