Skip to main content

Seminars

  • Better Web Security Through Adaptive Web Measurements

    PhD Thesis📅 08.09.2026 — 13:00 👤 Speaker: YIGIT SEVER 🎓 Supervisor(s): ASSOC.PROF.DR.PELIN ANGIN ULKUER 📍 Location: A105 ⏲ Duration: 120 min. 📝 Abstract: The Internet and the World Wide Web is a crucial service that we rely heavily on for our daily lives. Modern web applications carry immense importance through our reliance to…
  • COMMUNICATION-AWARE ACTIVESLAM FOR DECENTRALIZED SWARM EXPLORATION

    MSc Thesis📅 08.09.2026 — 10:00 👤 Speaker: IMRE KOSDIK 🎓 Supervisor(s): PROF.DR.ERTAN ONUR,ASSOC.PROF.DR.ELIF SURER 📍 Location: A105 ⏲ Duration: 60 min. 📝 Abstract: To create a map of unknown terrain where GPS is unreliable or unavailable, decentral ized UAV swarms are deployed. This task demands a trade-off between maximizing…
  • Cubic Bezier Splines on Polygon Meshes

    MSc Thesis📅 04.09.2026 — 15:00 👤 Speaker: SELAY TEKGUL 🎓 Supervisor(s): PROF.DR. YUSUF SAHILLIOGLU 📍 Location: A105 ⏲ Duration: 90 min. 📝 Abstract: This thesis develops a framework for constructing cubic Bézier splines on polygon meshes. Our method replaces the affine interpolation steps of de Casteljau's construction with…
  • INVESTIGATING OPTIMIZATION METHODS FOR SPECTRUM PRESERVATION IN SUBDIVISION SURFACES

    MSc Thesis📅 04.09.2026 — 14:00 👤 Speaker: FIRAT AGIS 🎓 Supervisor(s): PROF.DR.YUSUF SAHILLIOGLU 📍 Location: A105 ⏲ Duration: 90 min. 📝 Abstract: Subdivision schemes are widely used in digital geometry processing, but they are typically optimized for geometric continuity rather than spectral properties. In this thesis, we…
  • Transformer-Based Multi-Model Rank Fusion for Cross-Domain Recommendation and Dense Information Retrieval

    MSc Thesis📅 04.09.2026 — 11:30 👤 Speaker: MEHMET ERDENIZ AYDOGDU 🎓 Supervisor(s): PROF.DR.I.SENGOR ALTINGOVDE, PROF.DR.PINAR KARAGOZ 📍 Location: A105 ⏲ Duration: 90 min. 📝 Abstract: Rank fusion is a critical mechanism for synthesizing outputs from multiple retrieval models, yet traditional heuristic and linear supervised…
  • Towards Extreme Video Compression Using General World Knowledge

    MSc Thesis📅 03.09.2026 — 14:15 👤 Speaker: FURKAN KUCUK 🎓 Supervisor(s): ASSOC.PROF.DR.R.GOKBERK CINBIS 📍 Location: A105 ⏲ Duration: 90 min. 📝 Abstract: Classical and learned video codecs rebuild a generic visual prior at every receiver and transmit the residual it fails to predict. This thesis inverts that premise. A pretrained…
  • Generative Super-Resolution of 3D Gaussian Splat Sets

    MSc Thesis📅 03.09.2026 — 13:30 👤 Speaker: BATUHAN BAL 🎓 Supervisor(s): ASSOC.PROF.DR.R.GOKBERK CINBIS 📍 Location: A105 ⏲ Duration: 90 min. 📝 Abstract: 3D Gaussian Splatting (3DGS) enables real-time novel view synthesis, but Gaussian representations reconstructed from low-resolution observations are typically sparse and lack…
  • Random Prototype Knowledge Distillation for Cross-Architecture Knowledge Transfer

    MSc Thesis📅 02.09.2026 — 14:00 👤 Speaker: HAMZA ETCIBASI 🎓 Supervisor(s): ASSOC.PROF.DR.R.GOKBERK CINBIS 📍 Location: A105 ⏲ Duration: 90 min. 📝 Abstract: A widely held assumption in knowledge distillation is that teacher and student representations must occupy a compatible or explicitly aligned feature space. This thesis…
  • CONWISE-FT: ENHANCING FEW-SHOT LEARNING THROUGH CONTRASTIVE PROJECTION TUNING AND POST-TRAINING WEIGHT ENSEMBLING

    MSc Thesis📅 27.08.2026 — 15:00 👤 Speaker: ALPER BAHCEKAPILI 🎓 Supervisor(s): ASSOC.PROF.DR.EMRE AKBAS 📍 Location: A105 ⏲ Duration: 90 min. 📝 Abstract: Foundation vision–language models such as CLIP achieve strong zero-shot performance, but adapting them to specialised few-shot tasks still remains difficult. Two failure modes…
  • GRAPH-BASED CHAIN CONSTRUCTION APPROACH TO LANDMARK-GUIDED REINFORCEMENT LEARNING

    MSc Thesis📅 13.08.2026 — 15:00 👤 Speaker: ANIL ERDEM DERINOZ 🎓 Supervisor(s): PROF.DR.FARUK POLAT, ASST.PROF.DR.HUSEYIN AYDIN 📍 Location: A105 ⏲ Duration: 90 min. 📝 Abstract: Reinforcement Learning (RL) provides a powerful framework for autonomous deci-sion making; however, its real-world applicability is often bottlenecked by…