4. Vision research
cool products
https://youtu.be/hu4gMAvkIB4 (finger pointing translator)
https://bridge.occipital.com/
pose estimation using opencv
SOLVEPNP_ITERATIVE Iterative method is based on Levenberg-Marquardt optimization. In this case the function finds such a pose that minimizes reprojection error, that is the sum of squared distances between the observed projections imagePoints and the projected (using projectPoints ) objectPoints .
SOLVEPNP_P3P Method is based on the paper of X.S. Gao, X.-R. Hou, J. Tang, H.-F. Chang "Complete Solution Classification for the Perspective-Three-Point Problem". In this case the function requires exactly four object and image points.
SOLVEPNP_EPNP Method has been introduced by F.Moreno-Noguer, V.Lepetit and P.Fua in the paper "EPnP: Efficient Perspective-n-Point Camera Pose Estimation".
SOLVEPNP_DLS Method is based on the paper of Joel A. Hesch and Stergios I. Roumeliotis. "A Direct Least-Squares (DLS) Method for PnP".
SOLVEPNP_UPNP Method is based on the paper of A.Penate-Sanchez, J.Andrade-Cetto, F.Moreno-Noguer. "Exhaustive Linearization for Robust Camera Pose and Focal Length Estimation". In this case the function also estimates the parameters fx and fy assuming that both have the same value. Then the cameraMatrix is updated with the estimated focal length.
Four point and face
IMU and vision
IMU-AHRS library https://github.com/vshymanskyy/GY80
Sebastian O.H. Madgwick,"An efficient orientation filter for inertial and inertial/magnetic sensor arrays"
Fast IMU
http://www.analog.com/en/products/mems/accelerometers/adxl001.html#product-overview
Wearable Glass research
FPGA CNN
New xilinx product: The following links is the reVISION stack information:
https://www.xilinx.com/support/documentation/backgrounders/revision-overview.pdf
Video:
I am thinking that instead of buying the sdAccel development boards, is it more appropriate to buy the development kit that can run the reVISION stack apps?
The development kit that can run the reVISION stack apps:
https://www.xilinx.com/products/boards-and-kits/ek-u1-zcu102-es2-g.html
http://www.ece.cmu.edu/~coram/doku.php
http://darrinwillis.github.io/neuralHardware/
BTW Google already goes beyond FPGA for deep learning, they played the ASIC card:
Tracking
KCF paper: https://arxiv.org/abs/1404.7584
One-shot learning paper: https://arxiv.org/abs/1606.05233
arduino camera , face tracking arduino project
face detection demo https://youtu.be/JwX06vT4GOE
https://fyu.se/ 3d from a single view of a mobile phone.
4-point algo
demo https://www.youtube.com/watch?v=ZAewmLITfEM&feature=youtu.be
use 4-point to control a 3-D teapot https://youtu.be/3Ate7HD1pD8
SLAM
Slam: http://www.cvlibs.net/software/libviso/, http://www.cvlibs.net/publications/Kitt2010IV.pdf
The standard dataset KITTI : http://www.cvlibs.net/datasets/kitti/ video-demo
Deep net
Hintion
cuhk
jhu
Chinese words
http://www.cidianwang.com/shufa/
http://tool.wikichina.com/shufa/
http://humanum.arts.cuhk.edu.hk/Lexis/lexi-mf/search.php?word=武
summer research
http://calvinkam.science/vision
OCR
Compilation of Tesseract : http://calvinkam.science/vision/doku.php/tesseract
Tesseract C API: http://calvinkam.science/vision/doku.php/tesseract:api
Wearable glass projects
https://www.youtube.com/watch?v=ad2DNBWgB-0
3D
Tools :
An Open Source Feature Selection Repository in Python - scikit-feature http://www.kdnuggets.com/2016/03/scikit-feature-open-source-feature-selection-python.htm and http://www.public.asu.edu/~huanliu
Kinect
http://zugara.com/how-does-the-kinect-2-compare-to-the-kinect-1https://dev.windows.com/en-us/kinect/hardware
http://rpg.ifi.uzh.ch/research_dense.html
http://vision.in.tum.de/research/lsdslam
Labs:
Laboratory of Mathematical Methods of Image Processing headed by Prof. Andrey S. Krylov , http://imaging.cs.msu.ru/en
LSD-SLAM: Large-Scale Direct Monocular SLAM http://vision.in.tum.de/research/lsdslam
Tools:
Challenges