Training an artificial neural network with backpropagation algorithms requires an extensive computational process. Our recent work proposes to implement the backpropagation algorithm optically for in-situ training of both the linear and nonlinear diffractive optical neural networks which enables the acceleration of training speed and improvement on the energy efficiency on core computing modules. We numerically validated that the proposed in-situ optical learning architecture achieves comparable accuracy to the in-silico training with an electronic computer on the task of object classification and matrix-vector multiplication, which further allows adaptation to the system imperfections. Besides, the self-adaptive property of our approach facilitates the novel application of the network for all-optical imaging through scattering media. The proposed approach paves the way for the robust implementation of large-scale diffractive neural networks to perform distinctive tasks all-optically.
A novel video stitching algorithm is proposed which stitches video streams captured by multi-UAVs into panorama video streams. Each UAV captures one video streams, there are independent shakiness and parallax between each temporally aligned frames. A pure 2D homography does not have the capability to solve parallax like spatial-variant homographies, where each input video frame is sliced into grids and each grid is transformed with a local 2D homography. Video stitching has to consider video stabilization due to strong temporal correlations between frames. Video stitching and stabilization are combined, since they are both about putting the grids from video sources into panorama properly. Good stitching results are generated by optimizing the camera paths which describe the grid transformation between girds of frames before and after. In 4 aspects they are optimized: 1) spatially with the grids around, 2) temporally with the grid at the same place from before and after frames, 3) geometrically with the grids from another video, 4) reliably not too far from the original transformations. We focus on video stream processing which does not need the whole video frames, only 7 frames later than the input, state-of-the-art results are generated.
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.