3ea4597622
distances as similarities
55d5adbda9
Prototype distances/similarities map
9cf4fd47cd
Generate distances json
5ab0d57634
The data used in the notebook. Gathered from Atlas of Utopias
01b76e6d78
Python version for pyenv
61b61a0fbb
A notebook which explores mapping based on text similarity with Doc2Vec.
c4b7c28ca6
Sticky items for set range
a15195237d
drawing intermediate lines and playback icon
e25960f2e0
annotate.js dispatch pause event
c59b198b54
demo draw lines and click to playback
ef3fe001dc
override styles moved to collage
cd8a06a53b
list which processes occupy port
44a618a5ee
Refactor: prediction_server produces Frame objects
a3e42b4501
Try more detector options
f3b8e031c1
Fix dt
ec9bb357fd
Feature: render to zmq and tag frame with index
185962ace5
try ultralytics
081e26a458
try ultralytics
cc1e417db4
Try different trackers. Ultralytics & KeypointRCNN
2434470cdf
bypass renderer reads tracker data
994e150d37
start function
68bc273899
changes to make it run for me
a36db4e84b
note to future developers
d7b093b75b
Initial commit
06c5bde997
Option to save tracker output as training data
2171dd459a
switch to not predict video, but training data
f3ac903555
Auto-select latest iteration of model
0a02422733
allow disabling render preview
c903d07b49
Frame emitter loops over files in folder
4751d02229
Fix odd interval of waiting predictor
ec31810d80
Render directly to video
821d06c9cf
Test renderer to frames
3d34263a71
Gracefully stop all loops on crash of subprocess
9b39d7cd9b
Logging tweaks
27565d919e
Fix zmq-conflate and measure timings