Self‐Adaptive TiO 2 Optoelectronic Synapses via Microsecond–Nanosecond Pulse Modulation for Image Denoising
This work employs optical pulses spanning microsecond to nanosecond durations, demonstrating pulse‐length‐dependent photocurrent plasticity in a self‐adaptive manner. Learn‐forget cycles are realized through two distinct synaptic behaviors: depression and potentiation. Both are exploited for in‐sensor denoising of digital images, enhancing recognition accuracy from 92% to 96% and from 93% to 96%, respectively. This research provides an optoelectronic device scheme for high‐speed, low‐power neuromorphic vision systems.
Authors
- T. Wágner (ORCID: https://orcid.org/0000-0001-6445-5395)
- Guangyu Wen
- Petr Janíček (ORCID: https://orcid.org/0000-0002-2279-090X)
- Bo Zhang (ORCID: https://orcid.org/0000-0002-7863-9168)
- Xu Zhao
- Wenqiang Dong
Institutions
- University of Pardubice (CZ)
- Hebei Normal University (CN)
Publication Details
- Journal
- physica status solidi (a)
- Published
- 2026-09-01
- DOI
- https://doi.org/10.1002/pssa.70516
- Primary Topic
- Neural Networks and Reservoir Computing
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- Hebei Normal University