Commits

Branch selector

master

User selector

All users

All time

Commit history

Commits on Apr 28, 2020

• Adds default conversion from image to numpy during crowd dataset preprocessing

  golmschenk

  committed
  35b8075View commit details

  Copy full SHA for 35b8075

  Browse repository at this point

Commits on Aug 20, 2019

• Switch to age application as default

  golmschenk

  committed
  1058e74View commit details

  Copy full SHA for 1058e74

  Browse repository at this point

Commits on Aug 13, 2019

• Use true image patch indexes instead of random indexes for World Expo dataset

  golmschenk

  committed
  a02caecView commit details

  Copy full SHA for a02caec

  Browse repository at this point

Commits on Aug 12, 2019

• Fix World Expo dataset to work with the new network

  golmschenk

  committed
  b7acaddView commit details

  Copy full SHA for b7acadd

  Browse repository at this point

Commits on Aug 8, 2019

• Make the driving application the default, as it is the simplest

  golmschenk

  committed
  2a711b4View commit details

  Copy full SHA for 2a711b4

  Browse repository at this point

Commits on Aug 7, 2019

• Add map to the world expo dataset

  golmschenk

  committed
  566d05dView commit details

  Copy full SHA for 566d05d

  Browse repository at this point

Commits on May 30, 2019

• Fix bug where the database processors were using the word dataset instead of database

  golmschenk

  committed
  96f85d4View commit details

  Copy full SHA for 96f85d4

  Browse repository at this point

• Move database_directory to a property

  golmschenk

  committed
  a7d84f0View commit details

  Copy full SHA for a7d84f0

  Browse repository at this point

Commits on May 29, 2019

• Rename unlabeled/fake loss multiplier to the more correct matching/contrasting loss multiplier

  golmschenk

  committed
  f723c46View commit details

  Copy full SHA for f723c46

  Browse repository at this point

• Update default loss functions based analysis tests

  golmschenk

  committed
  2016364View commit details

  Copy full SHA for 2016364

  Browse repository at this point

Commits on May 17, 2019

• Rename the preprocessor to use database rather than dataset to be consistent

  golmschenk

  committed
  3ebc6b8View commit details

  Copy full SHA for 3ebc6b8

  Browse repository at this point

• Move all label generation code to the base dataset preprocessor

  golmschenk

  committed
  874bfcbView commit details

  Copy full SHA for 874bfcb

  Browse repository at this point

• Refactor the UCF CC 50 database to use the base class

  golmschenk

  committed
  0c3ae16View commit details

  Copy full SHA for 0c3ae16

  Browse repository at this point

• Refactor the UCF QNRF database to use the base class

  golmschenk

  committed
  e9d4068View commit details

  Copy full SHA for e9d4068

  Browse repository at this point

• Refactor the ShanghaiTech database to use the base level preprocessor

  golmschenk

  committed
  66e792bView commit details

  Copy full SHA for 66e792b

  Browse repository at this point

• Create base downloader and preprocessor

  golmschenk

  committed
  5e2334fView commit details

  Copy full SHA for 5e2334f

  Browse repository at this point

Commits on May 10, 2019

• Refactor run script to use formatted strings

  golmschenk

  committed
  e94876fView commit details

  Copy full SHA for e94876f

  Browse repository at this point

• Fix bug where model would start from the loaded model step when continue from existing is true, even though the directory did not already exist

  golmschenk

  committed
  6a831c2View commit details

  Copy full SHA for 6a831c2

  Browse repository at this point

Commits on May 9, 2019

• Replace the abstract inference method with a NotImplemented error, to allow the classes which don't implement it to still run

  golmschenk

  committed
  0387030View commit details

  Copy full SHA for 0387030

  Browse repository at this point

• Add the with_optimizers option to the DNN loader

  golmschenk

  committed
  23a2ea2View commit details

  Copy full SHA for 23a2ea2

  Browse repository at this point

• Move the inference functions to the the base srgan class

  golmschenk

  committed
  4c877abView commit details

  Copy full SHA for 4c877ab

  Browse repository at this point

Commits on May 7, 2019

• Remove the presentation and TensorBoard log to latex files. They have been moved to their own repository

  golmschenk

  committed
  ba47587View commit details

  Copy full SHA for ba47587

  Browse repository at this point

• Move coefficient presentation code within the coefficient module

  golmschenk

  committed
  d89857cView commit details

  Copy full SHA for d89857c

  Browse repository at this point

• Fix summary name for coefficient sorting

  golmschenk

  committed
  e78e88bView commit details

  Copy full SHA for e78e88b

  Browse repository at this point

Commits on May 6, 2019

• Add predicted labels to comparison values for output

  golmschenk

  committed
  3428926View commit details

  Copy full SHA for 3428926

  Browse repository at this point

• Refactor coefficient evaluation set and add RMSE summaries

  golmschenk

  committed
  608d6b9View commit details

  Copy full SHA for 608d6b9

  Browse repository at this point

• Allow for crowd inference

  golmschenk

  committed
  0459908View commit details

  Copy full SHA for 0459908

  Browse repository at this point

Commits on May 3, 2019

• Make UCF-QNRF unlabeled dataset start from end of labeled dataset

  golmschenk

  committed
  b97d814View commit details

  Copy full SHA for b97d814

  Browse repository at this point

Commits on Apr 15, 2019

• Add DenseNet model without map modules for crowd analysis

  golmschenk

  committed
  4b43e47View commit details

  Copy full SHA for 4b43e47

  Browse repository at this point

Commits on Apr 14, 2019

• Allow loading or continuing but not both at the same time. Also allow to load when continue is true, but continuation does not exist.

  golmschenk

  committed
  6f480c6View commit details

  Copy full SHA for 6f480c6

  Browse repository at this point

Commits on Apr 8, 2019

• Fix bug where the period model saving is using the wrong step save number

  golmschenk

  committed
  52b3f67View commit details

  Copy full SHA for 52b3f67

  Browse repository at this point

• Adding automatic saving on period

  golmschenk

  committed
  2a9c9eaView commit details

  Copy full SHA for 2a9c9ea

  Browse repository at this point

Commits on Apr 2, 2019

• Move the crowd dataset selection to an enum for easier selection

  golmschenk

  committed
  8a11784View commit details

  Copy full SHA for 8a11784

  Browse repository at this point

• Adds the crowd DGGAN experimental setup

  golmschenk

  committed
  5d231d5View commit details

  Copy full SHA for 5d231d5

  Browse repository at this point

Commits on Mar 20, 2019

• Fix bug where the wrong map name is used

  golmschenk

  committed
  565dcf5View commit details

  Copy full SHA for 565dcf5

  Browse repository at this point