~~

PhD "ZOOM" SEMINAR
Wednesday, June 24, 2020 at 14:00
Microknot optical resonators and their applications
Alexandra Blank
PhD of Dr. Yoav Linzon
Resonating microstructures defined locally on optical fiber tapers are an attractive venue for researchers in the recent years. These engineered microstructures find use in various applications, including photonic filters, lasers, optomechanical light-matter interaction devices, and various sensing elements mimicking human smell and taste sensing system (gas and liquids sensors), known as electronic noses and electronic tongues.
The presented work is devoted to experimental and numerical investigations of the behaviour of microknot resonators (MKRs) defined locally on optical fiber tapers as a promising platform for gas and liquid sensing applications. We demonstrate the two-probe localized heating technique for MKR fusing intended for improving its optical performance and mechanical stability in size and circular profile. We show that an above-threefold dynamical range enhancement has been consistently achieved. The fused MKRs are found to maintain phase stability, as opposed to the unfused knots exhibiting a random phase drift. We prove that e-fusing improves the mechanical strength in the MKR, providing it with transferability that is of importance in sensing applications.
We investigate fused MKRs’ characteristics for in-liquid sensing both local, where the analyte is delivered to the packaged device and the remote sensing. We have suggested and demonstrated two deployment schemes, namely, folded configuration intended for remote sensing, and straight configuration on either hydrophilic or hydrophobic substrate intended for packaged integration of photonic sensors. We found that in folded configuration reversibility remained sustainable over approximately 50 cycles of operation, whereas in most cases of the straight configuration on glass substrate transmission drops to the noise level after a few dewetting cycles. We show that a hydrophobic PDMS substrate can be advantageous relating to surface chemistry as compared to a typical hydrophilic glass substrate. For all those configurations we demonstrated their persistent sensing capabilities and defined unique recognition plots in principle components space specific to pure and diluted volatile liquids tested.
We demonstrate a durable and simple humidity sensing approach based on index-sensitive interference spectroscopy of surface stress birefringence and incorporating tapered microfibers on a silicon substrate coated with an active polymer layer. We show theoretically that the transmission spectrum of coated tapers possess interference patterns induced by the stress applied to the taper due to the coating weight load. The coated device demonstrated persistent detection capability in humid environment and a linear response to calibrated analytes. For each volatile organic analyte tested we defined a calibration plot in principal component space.
Microstructured tapers, when untreated, amenable to low performance in terms of resonance depth (dynamical range), Q and mechanical fragility. In this work we performed a novel comprehensive numerical study of the photonic transmission in manually prepared microknot resonators with different contact coupling area geometries and refractive index variations. Using selective modifications of the MKR coupling area and index profile, the transmission characteristics were studied, and a recipe for the experimental realization of high quality-factor resonators is prescribed.
We present results of the numerical study of the photonic transmission in 3D resonating microstructures defined on optical tapered fibers as functions of both radius and doping and deduce their sensitivities to localized refractive index variations in terms of resonance shifts. Based on the analysis of the transmission characteristics of polymer-coated microstructures, we demonstrated a linear response to the refractive index gradient in NIR wavelength range.
We envision that simple and robust devices with improved optical and mechanical characteristics will be harbored in next generation sensors, as well as optomechanical applications incorporating microfiber-based resonating structures.
Join Zoom Meeting
https://zoom.us/j/4962025174
The meeting will be recorded and made available on the School’s site.

~~

SCHOOL OF MECHANICAL ENGINEERING SEMINAR
Monday, April 27, 2020 at 14:00 
Wolfson Building of Mechanical Engineering, Room 206

Removal of acid dyes from concentrated synthetic and real dye effluent using biochar as sustainable adsorbent

Elizaveta Sterenzon
M.Sc. student of Prof. Hadas Mamane

Wastewater from the textile dyeing industry is a growing problem. It is often rich in color and dye residues, toxic compounds, chemicals, high BOD concentration and much harder degradation materials. The extended use of synthetic dyes, rather than the natural ones, makes the treatment of these textile effluents, much more difficult due to their high stability and low biodegradability. Different textile wastewater treatment technologies already exist in the market. However, in many cases, especially with household dyeing units, the textile wastewater is not treated and discharged directly into nearby water sources. There is a growing demand for low-cost, effective and sustainable alternatives for treatment methods, and adsorption occupies a prominent place as a method for dye removal. This study focused on the removal of acid dyes from synthetic and real dye effluent. For the synthetic effluent, Acid Violet 17 solution was prepared, and the real acid dye effluent was taken from dyer houses from Sirumugai, South India. Pine tree biochar was chosen as a low-cost, sustainable and innovative material for the adsorption. Due to its high porosity and large surface area, biochar is gaining a lot of interest in the field of water treatment, as a low-cost contamination adsorbent. Various parameters were tested to investigate the dye removal ability and the best adsorption conditions, as temperature, solution pH and biochar dosage. Kinetic and thermodynamic studies also have been conducted. The results showed an increased dye adsorption at lower pH and at higher temperatures, which indicates an endothermic reaction. According to the kinetic studies, the governing model for the adsorption is the pseudo second order and the Langmuir isotherm fitted significantly better than Freundlich. Regeneration of the biochar was found to be very effective with absolute ethanol. After 5 regeneration cycles the adsorption ability reduced by only 3% from the initial dye removal, and it was almost no biochar mass loss observed. The real textile effluent demonstrated high adsorption affinity to the biochar as well, with almost 100% dye removal after 60 min of treatment.

השתתפות בסמינר תיתן קרדיט שמיעה = עפ"י רישום שם מלא + מספר ת.ז.  בצ'אט

Join Zoom Meeting
https://us04web.zoom.us/j/6207215556

Speaker: Maayan Shuvi

M.Sc. student under the supervision of Prof. Daniel Cohen-Or

Wednesday, April 22^nd, 2020 at 14:00

     ZOOM

Neural Alignment for Face De-pixelization

Abstract

We present a method to reconstruct a high-resolution video from a face-video, where a person’s identity was obscured by pixelization. This hiding method is popular because the viewer can still perceive a human face figure and the overall head motion. However, we show in our experiments how a fairly good approximation of the original video is reconstructed so that anonymity is compromised.

Our system exploits the simultaneous similarity and small disparity between close-by video frames depicting a human face, and employs a spatial transformation component that learns the alignment between the pixelated frames.

Each frame, supported by its aligned surrounding frames, is first encoded, then decoded to a higher resolution. Reconstruction and perceptual losses promote adherence to the ground-truth, and an adversarial loss assists in maintaining domain faithfulness. There is no need for explicit temporal coherency loss as it is maintained implicitly by the alignment of neighboring frames and reconstruction. 

Our work shows that given the specific prior of a human face, which is well structured, and using multiple aligned frames, there is enough information in a pixelated video stream to reconstruct a high-quality approximation of the original signal.