Face is an important source of information for interpersonal communication and recognition, so facial reconstruction technology has always been a research focus in the field of computer vision. We present a deep learning-based algorithm for accurate three-dimensional face reconstruction from a single-view video stream. The proposed method processes the input video stream frame by frame, extracts facial region information, and constructs a facial reconstruction network that utilizes 3D Morphable Models to reconstruct the precise geometric shape of the face. Additionally, we design a multitiered loss function, including low-level pixel consistency loss, facial landmark loss, and high-level identity loss. Furthermore, these multi-tiered losses are utilized as weak supervision signals to guide the supervised learning of the reconstructed face, thereby enhancing the quality and accuracy of the reconstruction.
We investigate on a multi-wavelength operation erbium-doped fiber laser Q-switched by a graphene-based saturable absorber. Stable pulses were generated with the widths from 6.9 to 1.5 μs, energies from 40.4 to 130.2 nJ and repetition rates from 68.32 to 132.9 kHz, when the pump power increased from 142.32 to 441.86 mW. A fiber Bragg grating with five reflective peaks was inserted into the fiber ring through an optical circulator, resulting in a stable output of five-lasing-wavelength output. The laser can perform as a low-cost and easy-built all fiber light source, and has potential applications in the fields where pulses at multi-wavelength operation are needed, i.e., temperature or strain fiber sensors.
It is very important to understand the characteristics of gain saturation for homogeneously broadened laser medium. Assuming a Lorentzian lineshape function with a linewidth of ΔνH, we have derived the analytical functions of the gain coefficient under three different conditions: 1) a small and 2) a large signal incident on the medium respectively, and 3) a small signal accompanied by a large signal simultaneously incident on the medium. We have found that the bandwidth Δν of the gain coefficient is equal to the linewidth ΔνH for condition 1) and 3) while it is Δν = √1+Iν1 / IsΔνH for condition 2). Here, Iν1 and Is are the intensity of the large signal and the saturation intensity at the center frequency ν0, respectively. The reasons are also presented for such results. For condition 2), gain saturation effect is strongly dependent on the frequency of the large signal: the more the frequency deviates from the center frequency, the weaker the gain saturation effect. This results an increase of the bandwidth. For condition 3), the intensity of the large signal only changes the distribution of populations on the upper and lower energy levels. Hence, the shape of the gain coefficient does not change with the gain reduction of the small signal, thus, its bandwidth remains the same.
Experimental studies on the mode competition in CW fiber laser pumped Raman fiber lasers with two different F-P cavities are carried out. The first cavity consists of a dichroic mirror and a cleaved fiber end and the second of two dichroic mirrors. Results show that there is drastic competition in the first cavity and much less one in the second because of efficient lasing mode selection of two dichroic mirrors. Reasons for mode competition are analyzed. There are two main kinds of competition in the cavity: one exists between the 440cm-1 peak and the 490cm-1 peak in Raman scattering spectrum; and the other is related with long cavity and spatial hole burning. Output characteristic of the laser and probability of different mode wavelengths are also measured.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.