{"686091":{"#nid":"686091","#data":{"type":"event","title":"Ph.D. Dissertation Defense - Pin-Jui Ku","body":[{"value":"\u003Cp\u003E\u003Cstrong\u003ETitle\u003C\/strong\u003E\u003Cem\u003E:\u0026nbsp; Incorporating Geometric and Consistency Constraints with Deep Models for Robust Phase Reconstruction and Speech Enhancement\u003C\/em\u003E\u003C\/p\u003E\u003Cp\u003E\u003Cstrong\u003ECommittee:\u003C\/strong\u003E\u003C\/p\u003E\u003Cp\u003EDr.\u0026nbsp;Chin-Hui Lee, ECE, Chair, Advisor\u003C\/p\u003E\u003Cp\u003EDr.\u0026nbsp;Larry Heck, ECE\u003C\/p\u003E\u003Cp\u003EDr.\u0026nbsp;David Anderson, ECE\u003C\/p\u003E\u003Cp\u003EDr.\u0026nbsp;Elliot Moore, ECE\u003C\/p\u003E\u003Cp\u003EDr.\u0026nbsp;Marco Sabato Siniscalchi, U of Palermo\u003C\/p\u003E","summary":"","format":"limited_html"}],"field_subtitle":"","field_summary":[{"value":"\u003Cp\u003EThis dissertation proposes a new framework for phase estimation that overcomes these limitations and demonstrates its effectiveness across multiple DNN-based SE models. We begin by introducing the first deep state-space-based SE model operating on complex-valued spectrograms. While it surpasses baseline models with a compact U-Net architecture, its estimated phase offers limited improvement over the noisy phase, underscoring the difficulty of direct phase prediction. To address this, we develop a novel explicity consistency-preserving loss that leverages the observation that perceptually high-quality speech arises when magnitude and phase are mutually consistent. Building on this insight, we integrate geometric constraints under additive noise conditions with the consistency principle, resulting in the Multi-Sourced Griffin-Lim Algorithm (MSGLA). MSGLA jointly refines speech and noise phases through iterative updates guided by DNN-estimated magnitudes and geometric relationships, outperforming direct phase estimation and prior geometric methods. Finally, we extend these ideas to a large-scale generative pretraining framework that models the distribution of clean speech spectrograms and incorporates the consistency-based phase loss during training. \u0026nbsp;\u003C\/p\u003E","format":"limited_html"}],"field_summary_sentence":[{"value":"Incorporating Geometric and Consistency Constraints with Deep Models for Robust Phase Reconstruction and Speech Enhancement "}],"uid":"28475","created_gmt":"2025-10-29 20:15:39","changed_gmt":"2025-10-29 20:17:01","author":"Daniela Staiculescu","boilerplate_text":"","field_publication":"","field_article_url":"","field_event_time":{"event_time_start":"2025-11-19T08:30:00-05:00","event_time_end":"2025-11-19T10:30:00-05:00","event_time_end_last":"2025-11-19T10:30:00-05:00","gmt_time_start":"2025-11-19 13:30:00","gmt_time_end":"2025-11-19 15:30:00","gmt_time_end_last":"2025-11-19 15:30:00","rrule":null,"timezone":"America\/New_York"},"location":"Room 530, TSRB","extras":[],"groups":[{"id":"434381","name":"ECE Ph.D. Dissertation Defenses"}],"categories":[],"keywords":[{"id":"100811","name":"Phd Defense"},{"id":"1808","name":"graduate students"}],"core_research_areas":[],"news_room_topics":[],"event_categories":[{"id":"1788","name":"Other\/Miscellaneous"}],"invited_audience":[{"id":"78771","name":"Public"}],"affiliations":[],"classification":[],"areas_of_expertise":[],"news_and_recent_appearances":[],"phone":[],"contact":[],"email":[],"slides":[],"orientation":[],"userdata":""}}}