JCJoseph Clements

Joseph Clements

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Joseph Clements has a total of 11 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Thermoresponsive alginate/PNIPAM/Fe3O4 microbots under magnetic surface locomotion for ocular therapies

    by Amira Khedewy, Lasheen Gurusinghe, Elizabeth Fox, Sangwon Lee, Joseph Clements, Yasin Cagatay Duygu, Min Jun Kim, Louis William Rogowski


  • Open Access

    Thermoresponsive Alginate/PNIPAM/Fe3O4 Microbots under Magnetic Surface Locomotion for Ocular Therapies

    by Amira Khedewy, Lasheen Gurusinghe, Elizabeth Fox, Sangwon Lee, Joseph Clements, Yasin Cagatay Duygu, Min Jun Kim, Louis Rogowski

    Magnetically actuated microbots capable of controlled locomotion and material adaptability are of growing interest for minimally invasive biomedical applications, particularly in confined and soft biological environments. This work presents the design, fabrication, and magnetic actuation of thermoresponsive alginate/poly(N-isopropylacrylamide) (PNIPAM) hydrogel microbots for future ocular therapies. The composite hydrogel exhibits a reversible lower critical solution temperature (LCST) transition between 28–40 °C, producing temperature-dependent changes in viscosity and optical appearance. Spherical microbots (average diameter ~194 µm) were fabricated using a centrifugation-assisted extrusion method and uniformly loaded with iron oxide nanoparticles to enable untethered magnetic actuation. Under a 10 mT rotating magnetic field, the microbots demonstrated stable rolling locomotion with a linear velocity–frequency relationship between 0.25–2 Hz, achieving slip ratios near unity (1.03 ± 0.09). Open-loop, closed-loop, and coordinated group actuation yielded consistent translational velocities ranging from 112–169 µm/s. Group-level actuation showed collective trajectory following with average heading errors of ~12.4° and lateral errors of ~17 µm, while closed-loop control enabled convergence to sequential targets with steady-state errors of ~45 µm (~23% of microbot diameter). As a proof of concept for non-planar environments, microbots successfully climbed curved, ocular-inspired PDMS surfaces up to ~50° These results establish alginate/PNIPAM hydrogels as a viable material platform for responsive microrobotic systems and provide the first demonstration of controlled magnetic locomotion in a hydrogel microbot capable of temperature-dependent material property switching. The combined actuation and thermoresponsive behavior highlight potential utility in minimally invasive biomedical applications, including targeted therapy, ocular microsurgery, and particularly, future use in military-relevant injuries such as vitreous hemorrhage recovery.

  • Reliable Hardware Watermarks for Deep Learning Systems

    by Joseph Franklin Clements, Yingjie Lao

    Recent successes in deep learning have indicated that hardware technologies will play a prominent role in future deep learning industries and applications. In light of their value, researchers have recognized that deep neural networks (DNNs) and other deep learning intellectual properties (IPs) can be easily pirated, especially in undefended settings. While multiple avenues of defending deep learning systems have been identified, watermarks are particularly valuable as they allow IP theft to be identified and remedied when it occurs. However, such defenses have yet to be considered for defending the hardware platforms running the deep learning systems. This article presents the first framework for applying watermarks toward defending deep-learning hardware accelerators from piracy, called DeepHardMark. The proposed methodology embeds modifications into the functional blocks of deep-learning hardware accelerators to act as a watermark signature. These modifications produce targeted alterations to the execution of key DNNs on corresponding key samples, which identifies the hardware. We optimize this methodology to simultaneously minimize the impact of the watermark embedding on both the hardware and algorithmic components of the deep learning system making the watermark unobtrusive and challenging to detect. Our experimental evaluations demonstrate the feasibility of embedding the proposed modifications into typical hardware designs and in various deep-learning scenarios.