Progress for week 42 (2015)

Fra Robin

(Forskjeller mellom versjoner)
Gå til: navigasjon, søk
 
(14 mellomrevisjoner ikke vist.)
Linje 14: Linje 14:
* Implement ZMP calculator on NAO robot
* Implement ZMP calculator on NAO robot
 +
== [[user:perwn|Per Wessel]] ==
 +
=== Budget ===
 +
* Plan data gathering
 +
=== Accounting ===
 +
* Not done, meeting with CTO and CEO of chemisense today
 +
 +
== [[user:Eirisu|Eirik Sundet]] ==
-
== Magnus Olden ==
 
=== Budget ===
=== Budget ===
-
* Get all data needed
+
* Implement C++ version of Road Segmentation using RGB-space for Gaussian modelling, for single images
-
* Structure data
+
* Implement update function for the Gaussian model
-
* Read about fuzzy neural networks
+
* Implement Mixture-of-Gaussian classifier
 +
* Update implementation for video compatibility.
 +
* Test parameters for processing speed and accuracy
=== Accounting ===
=== Accounting ===
-
* Gotten all data
+
* 1: Done
-
* Started structuring
+
* 2: In progress
-
* Read articles about fuzzy NN
+
* 3: In progress
 +
* 4: Not started
 +
* 5: Not started
-
== Karl Magnus ==
+
== [[user:vegardli|Vegard]] ==
=== Budget ===
=== Budget ===
-
* Find good dataset to test with
+
* Tweak GA for better results
-
* Test CNN and brute force matching
+
* Start work on CUDA implementation
=== Accounting ===
=== Accounting ===
-
* Done
+
* Spent more time than expected on bugfixing and optimization of application/CPU implementation
-
* Did worse then direct image ADS
+

Nåværende revisjon fra 22. okt 2015 kl. 12:04

Innhold

Student template (copy this for your entry)

Budget

  • Todo 1
  • Todo 2

Accounting

  • Done 1
  • Done 2

Tomas Renaa

Budget

  • Check reliability of NAOs feet sensors, feedback for phase resetting.
  • Check reliability of gyroscopic sensor, can it be used for feedback to CPG-network?
  • Implement ZMP calculator on NAO robot

Per Wessel

Budget

  • Plan data gathering

Accounting

  • Not done, meeting with CTO and CEO of chemisense today

Eirik Sundet

Budget

  • Implement C++ version of Road Segmentation using RGB-space for Gaussian modelling, for single images
  • Implement update function for the Gaussian model
  • Implement Mixture-of-Gaussian classifier
  • Update implementation for video compatibility.
  • Test parameters for processing speed and accuracy

Accounting

  • 1: Done
  • 2: In progress
  • 3: In progress
  • 4: Not started
  • 5: Not started

Vegard

Budget

  • Tweak GA for better results
  • Start work on CUDA implementation

Accounting

  • Spent more time than expected on bugfixing and optimization of application/CPU implementation
Personlige verktøy