Skip to content
GitLab
Explore
Sign in
Primary navigation
Search or go to…
Project
I
ifind1_scanplanes_tensorflow
Manage
Activity
Members
Labels
Plan
Issues
Issue boards
Milestones
Wiki
Code
Merge requests
Repository
Branches
Commits
Tags
Repository graph
Compare revisions
Snippets
Build
Pipelines
Jobs
Pipeline schedules
Artifacts
Deploy
Releases
Container Registry
Model registry
Operate
Environments
Monitor
Incidents
Service Desk
Analyze
Value stream analytics
Contributor analytics
CI/CD analytics
Repository analytics
Model experiments
Help
Help
Support
GitLab documentation
Compare GitLab plans
Community forum
Contribute to GitLab
Provide feedback
Keyboard shortcuts
?
Snippets
Groups
Projects
Show more breadcrumbs
Bernhard Kainz
ifind1_scanplanes_tensorflow
Graph
4a5372813646537a60c31cf6fe5191a4088ad5a5
Select Git revision
Branches
3
biggerdata
master
default
normalise
3 results
You can move around the graph by using the arrow keys.
Begin with the selected commit
Created with Raphaël 2.2.0
19
Jun
27
May
24
21
20
19
18
12
11
7
2
30
Apr
27
24
23
22
13
1
Mar
25
Feb
24
18
17
16
18
Jan
fixed bugs and added a lot of results
biggerdata
biggerdata
fixed data augmentation bugs
switched to processing things with numpy, very fast whilst threading
asynchronous is still quite slow increased thread count to 2
added asynchronous prefetching of data otherwise training was simply too slow
fixed temporal model over videos and should have fixed temporal restoring
adding plotting for the normal convnets and have evaluation over long videos working but not classifying very well
Added graphing capabilities and almost final networks, still large errors though
fixed formatting bug, changed temporal model
All networks should technically be functionin. Seeign some strange results on temporal. Tidied up code
GCNN should be conneted to temporal model for training
Added loading of vidlets to be trained, and fixed the original network
resliased numpy could not deal with such large arrays so implemented the random batching, currently running now
removed prints and fixed network structure
seems to able to train on the TCNN, now need to get the evaluation working, and fix restoring the model
basics of the network should be ok, now need to make it run properly
Annoying edge case in getting batch data fixed
Normal network should be working, with starts on the temporal model. Fixed a bug in classification report
Includes the working benchmark, and should be able to process and amount of data
altered master to only have 14 labels, but failure on saving
master
master
changed folder path for data
tried to fiddle with array types to fix training, but to no avail yet
The big data is successfully recombined into one big numpy array however cuda says out of memory
I have split the datasets using h5py overcoming numpys problem of having too much data.
added label manager to tidy up code
normalise
normalise
changing the code to process the larger data, cannot save the train data, as it is too large, this is a known problem in python 2.7, I will fix this in the next commit
Added code for preprocessing large iFind dataset.
More features in 2nd layer I believe suffered from overfitting, reverting change
changed the number of features and size of kernels
Removed dropout in training and received much better results
Removed .npy files
Removed compiled files and .npy files
Tried Adding tensorboard to visualise the network. Also added batch normalisation
initial commit
Loading