asteroza + vision   139

Evolving Images for Visual Neurons Using a Deep Generative Network Reveals Coding Principles and Neuronal Preferences: Cell
Using some specific monkey neurons that are in a cluster near the ears as a replacement for the discriminator in a GAN deep learning setup, these biology/neurology researchers effectively were able to ask the cells directly what visually interested them.
biology  neurology  monkey  brain  GAN  dep  machine  learning  visual  cortex  facial  recognition  vision  perception 
may 2019 by asteroza
Good Strategy/Bad Strategy | The Difference and Why It Matters
Apparently a pretty good book. Clearly outlines why most strategies are not real strategies and/or they suck.
business  management  productivity  book  reference  information  strategy  startup  vision 
april 2019 by asteroza
Towards Robotic Feeding: Role of Haptics in Fork-Based Food Manipulation - IEEE Journals & Magazine
Two stage neural net, using a general object classifier to locate food on a plate, then a second stage to ID the food to match against a database of force factors for how a robot arm holding a fork should skewer the food to grab it.
deep  machine  learning  vision  computer  neural  network  RetinaNet  SPnet  algorith  robotics  robot  arm  human  assistance  assistive  medicine  health  feeding 
march 2019 by asteroza
The Sound of Pixels
Recreating sound from images/video alone, sorta like lip reading. But it actually seems to pull off making music again (I suppose hirez video might be necessary)
deep  machine  learning  computer  vision  audio  sound  emulation  simulation  replication  music 
november 2018 by asteroza
PixFood - Take a Photo! Get Tasty Recipes!
Commercializing NotHotdog to be a recipe recommender based on viewable ingredients (type and size/quantity estimation?)
iPhone  iOS  android  app  software  computer  machine  vision  recipe  recommendation 
november 2018 by asteroza
NotBacon | Microsoft Docs
So Microsoft/Azure felt left out after all the HotdogOrNot hype so...

But this is mostly a demo for Azure cloud hosted machine vision, not like HotdogOrNot's full offline neural net image recognition.
microsoft  azure  deep  machine  learning  computer  vision  bacon  C#  javascript  cloud  service 
october 2018 by asteroza
Neural scene representation and rendering | Science
Google using image sensor data to make 3D models from 2D pictures, using a GQN (not a GAN)
google  AI  machine  deep  learning  GQN  artificial  vision  image  processing  2D  3D 
june 2018 by asteroza
Facial recognition used to analyze students' classroom behaviors - People's Daily Online
So, if this was genuinely detecting disinterest and confusion emotions, then this would be a effective means of improving education methods and teacher styles via direct feedback, but since this is also being used for surveillance as well...
china  AI  facial  expression  machine  computer  vision  emotion  detection  recognition  education  learning  school  management  surveillance 
may 2018 by asteroza
Seek App ·
Using machine learning to identify animals and plants. Probably shouldn't use it for mushroom differentiation quite yet...
iPhone  android  app  software  machine  vision  image  recogniton  deep  learning  anima  plant  photo  species  identification  observation 
march 2018 by asteroza
[1712.09665] Adversarial Patch
Designing adversarial sticker patches to mess with machine/computer vision systems
adversarial  patch  sticker  camouflage  machine  learning  attack  security  research  computer  vision 
january 2018 by asteroza
Near term they are offering vision systems for product defect detection
AI  manufacturing  product  defect  machine  computer  vision  inspection 
december 2017 by asteroza
Revolutinary social video platform
Using AI to discover context/objects in video and surfacing it to users in a targeted way, rather than YouTube single frame previews in the slider/scrubber.
machine  learning  video  object  context  recognition  computer  vision  advertising 
november 2017 by asteroza
Not Hotdog - Android Apps on Google Play
Not Hotdog is an app that tells you whether your food is a hotdog or not hotdog.
not  hotdog  android  app  software  computer  machine  vision  deep  learning  AI 
july 2017 by asteroza
How HBO’s Silicon Valley built “Not Hotdog” with mobile TensorFlow, Keras & React Native – Medium
How Not Notdog the app was built. Lots of interesting tricks, though some of the practical tuning tips may be dependent on the small neural network size and focus on recognizing one thing...
humor  HBO  silicon  valley  not  hotdog  app  software  iOS  iPhone  android  machine  vision  deep  leaning  computer  mobile  TensorFlow  Keras  react.js  nerual  network  training 
june 2017 by asteroza
Cut down QRcodes suitable for outdoor environments being viewed from extreme distance/angles for better machine vision recognition.
AprilTag  long  distance  machine  computer  vision  object  tracking  2D  barcode  QRcode  recognition  unmanned  vehicle  navigation  UAV  UGV  Delicious 
march 2017 by asteroza
ageitgey/show-facebook-computer-vision-tags: A very simple Chrome Extension that...
Firefox,chrome plugin that demos Facebook's computer vision API for object recognition to tag stuff in images.
chrome  firefox  addon  plugin  extension  facebook  computer  vision  object  recognition  API  demo  PoC  Delicious 
january 2017 by asteroza
Movidius | Visual Sensing for the Internet of Things
So they have a USB dongle version usable for UAV drone SLAM navigation apparently...
neural  network  machine  computer  vision  chip  SoC  hardware  electronics  devices  Delicious 
may 2016 by asteroza
Vision API - Image Content Analysis — Google Cloud Platform
google  cloud  machine  vision  API  object  classification  recognition  service  Delicious 
december 2015 by asteroza
Polarized 3D - Achuta Kadambi - MIT Media Lab
Uses two 3D imaging techniques to offset the negatives of each technique, depth field and obliqueness, to gain increased resolution.
polarized  light  depth  field  high  resolution  3D  scanning  algorithm  machine  computer  vision  research  technology  Delicious 
december 2015 by asteroza
Object recognition for robots
SLAM algorithms may be better for object recognition?
MIT  SLAM  algorithm  machine  object  vision  software  research  Delicious 
august 2015 by asteroza
Inceptionism: Going deeper into Neural Networks
Google is feeding neural networks pictures that resemble machine hallucinations. This may not end well...
neural  network  machine  vision  object  recognition  tagging  research  google  hallucinationtraing  art  picture  image  photo  Delicious 
june 2015 by asteroza
welcome [Rapyuta]
Robot intelligence offload to the cloud, so the robot become more of a remote sensor data collection and transfer platform with servo microcontrollers, and the actual machine vision or motion planning is done in the cloud. Makes the robots cheaper, the machine vision faster, and can use a larger object recognition library and motion library.
Rapyuta  cloud  robotics  framework  service  platform  intelligence  offload  ROS  machine  vision  object  recognition  sensor  data  processing  Delicious 
june 2015 by asteroza
ZenRobotics - Intelligent Robotic Systems
An AI trash picker, to pick out recyclables from trash to make a profit.
robotics  picker  trash  garbage  recycling  sorter  hardware  electronics  devices  AI  machine  computer  vision  Delicious 
june 2015 by asteroza
EnChroma | Color For The Color Blind
Uses an optics trick originally used for laser protection goggles for laser surgery, it filters out precise wavelengths (yellow notch filter) that cause simultaneous red/green cone stimulation in colorblind individuals with that type, leaving the rest to be properly sensed by the respective cones.
eyeware  eyewear  glasses  sunglasses  spectrum  filter  medicine  health  colorblind  colorblindness  vision  hardware  spectral  notch  multinotch  Delicious 
march 2015 by asteroza
Skybox likely used this middleware code to link openCV and other image object recognition software to bulk process satellite imagery for their service, as they originally had been using Hadoop to process their imagery.
MR4C  mapreduce  C  C++  native  code  execution  hadoop  google  skybox  opensource  software  remote  sensing  imagery  image  picture  photo  machine  vision  object  recognition  bulk  processing  aerial  surveillance  satellite  geospatial  Delicious 
march 2015 by asteroza
Trying to improve computer vision on drones for GPS denied environments and general obstacle avoidance.
UAV  drone  machine  computer  vision  navigiation  Delicious 
january 2015 by asteroza
CVPR 2014 3D Reconstruction from Accidental Motion
So sorta Lytro, in that it's basically interpolating to build the fake depth, and could get tricked by translucent/reflective surfaces or unexpected motion. Some of the tricks are apparently going into google's camera app for android 4.4 allegedly.
optics  computer  vision  camera  software  3D  reconstruction  shallow  depth  field  Delicious 
april 2014 by asteroza
Visual Media Reasoning (VMR)
VMR components are scheduled to be open sourced at the DARPA Open Source catalog soon.
VMR  visual  media  reasoning  image  picture  photo  video  object  recognition  mapping  search  query  software  research  technology  DARPA  intelligence  military  analytics  opensource  machine  vision  computer  Delicious 
february 2014 by asteroza
A feature construction method for general object recognition
Apparently a new high performance object recognition algorithm, but not a normally derived genetic algorithm.
machine  vision  object  recognition  genetic  algorithm  evolution  constructed  software  Delicious 
january 2014 by asteroza
« earlier      
per page:    204080120160

related tags

2.0  2D  3D  802.15.4a  academic  accelerated  accelerator  accessory  active  ADAS  addon  adjustable  adversarial  advertising  aerial  agriculture  AI  ALE  algorith  algorithm  algorithms  alpha  AlphaMicron  amazon  amplification  analysis  analytics  android  anemia  anima  anonymity  anti-lamenessing  anti-OCR  anti-surveillance  antispam  antisurveillance  API  app  application  AprilTag  AR  architecture  ArduEye  arduino  arm  armored  array  art  artificial  ARToolkit  assist  assistance  assistant  assisted  assistive  attachment  attack  audio  augmentation  augmented  authentication  automatic  automation  automotive  autonomous  autonomy  AutoSmiley  autostabilization  avatar  avoidance  awareness  azure  baby  bacon  barcode  bed  behavior  bifocal  big  bigdata  biology  biometrics  biomimicry  blind  board  book  brain  brand  bulk  business  C  C#  C++  C71  cadmium  camera  camouflage  cancer  captcha  capture  car  card  cat  categorization  cell  cellphone  center  challenge  character  china  chip  chrome  classification  clipart  cloud  code  color  colorblind  colorblindness  colorized  comparison  compound  compputer  compund  computational  computer  computing  constructed  construction  contact  contactless  context  contrast  control  Cornell  corner  correction  cortex  cost  countersurveillance  crystal  CUDA  curved  custom  CV  cybernetics  DARPA  data  decode  decor  deep  defect  defense  Delicious  demo  dense  dep  depth  design  detection  development  devices  digital  discrimination  display  distance  door  doorknob  driver  driving  drone  dystopia  e-ink  e-paper  e-tint  editing  education  electronics  embedded  emoticon  emotion  emulation  energy  engine  enhancement  environment  ER  eulerian  europe  evesdropping  evolution  execution  expanded  expression  extension  eye  eyeball  eyeglasses  eyesight  eyetap  eyeware  eyewear  face  facebook  facial  factor  fast  fature  feedback  feeding  field  film  filter  fingernail  firefighter  firefox  flash  flexible  FLIR  flow  focus  font  forehead  form  fovea  framerate  framework  FTPE  future  GAN  gap  garbage  gaze  GazoPa  gear  generator  genetic  geoanalytics  geolocation  geometric  geospatial  geotagging  gesture  GigE  GIS  glass  glasses  globe  google  GoPano  GPGPU  GPU  GQN  graphic  graphics  green  hacking  hadoop  hallucinationtraing  hand  handicapped  handle  haptic  hardawre  hardware  harvester  HBO  HCI  head  headset  health  helmet  hidden  hierarchical  high  HMD  holography  hotdog  HTM  HUD  human  humor  icon  identification  illusion  image  imagery  images  IMAGINATION  imaging  impaired  index  industrial  inertial  information  infrared  Innovega  input  INS  insect  inspection  integrated  intel  intelligence  interface  interior  internet  iOptik  iOS  iPhone  IR  ISP  japan  java  javascript  KDDI  Keras  kickstarter  Kinect  language  laser  Laura  LCD  leaning  learning  LED  lens  lettuce  library  light  lightguide  link  liquid  list  localisation  logo  long  low  mac  machine  magnification  Mali  management  manufacturing  map  mapping  mapreduce  materials  media  mediated  medicine  microphone  microscope  microsoft  military  MIT  mobile  model  modeling  modification  monitor  monitoring  monkey  motion  mounted  MR  MR4C  multinotch  music  nanophotography  native  navigation  navigiation  neocortex  nerual  network  neural  neurology  Neuroph  neuroscience  ngine  night  NIR  norway  not  notch  novel  Numenta  NuPIC  Nvidia  obfuscation  object  observation  obstacle  OCR  oculus  offensive  offload  omnidirectional  omnifocal  omnigraffle  Omron  online  OpenCL  openCV  OpenCV  openGL  OpenGL  OpenNI  opensource  OpenVIDIA  operational  OPSEC  optical  optics  OrCam  OSX  outdoor  overlay  panaoramic  PANOPTES  panorama  panoramic  paper  parallel  patch  pattern  PCIe  PCIexpress  peephole  penetrating  perception  performance  peripheral  personal  phone  photo  PhotoContext  photography  photos  picker  picture  pipeline  plan  planar  plant  platform  plugin  PMcode  PoC  polarized  PolarRose  polymer  portable  postprocessing  power  PR  prism  privacy  PrivateEye  processing  processor  product  productivity  ProFORMA  programming  project  prosthetic  protection  protective  prototype  psychology  Pythia  python  PyTorch  QRcode  Qualcomm  query  radio  ranging  Rapyuta  ray  react.js  reading  reality  realtime  reasoning  reCAPTCHA  recipe  recognition  recogniton  recommendation  recongition  reconstruction  recycling  reference  reflection  refractive  remote  rendering  replacement  replication  research  resolution  retina  RetinaNet  rift  robot  robotics  rockwell  rollable  room  ROS  SAMWISE  satellite  scanner  scanning  scene  school  science  screen  search  security  seethrough  segmented  self  semantic  Sense  sensing  sensor  service  sewing  shadertoy  shadow  shallow  shouldersurfing  signal  silicon  similar  simulation  simultaneous  size  skybox  Skynet  SLAM  sleep  small  smartphone  smile  smiley  smilie  SNS  SoC  social  software  solar  sorter  sound  space  spam  spatial  species  spectra  spectral  spectrum  speech  SPnet  SR  stabilization  standard  startup  stencils  sticker  stitching  strategy  structure  sunglasses  Sunovia  super-resolution  Superfocus  surveillance  survival  sweden  synthetic  system  tactical  tactile  tagging  tank  technology  telepresence  telluride  template  templates  temporal  tensorflow  teradeep  terminal  test  testing  tetrachromacy  text-to-speech  texture  time-of-flight  tint  tomography  toolkit  tools  tracking  training  transparent  trash  TTS  turing  typeface  UAV  UAVE  UGV  UI  unmanned  Unreal  upload  urban  usage  use  utilities  UV  valley  variable  variance  vehicle  video  videoconferencing  ViPR  virtual  visibility  visible  vision  visual  visualization  VMR  voice  VR  vRD  wall  wearable  web  webAPI  webcam  webcomic  webGL  wifi  windows  wireless  wolfram  writing  xray  YOLO  zigbee  ZXX 

Copy this bookmark: