Making the best of poor video quality

Molly Roberts

unread,

Jun 25, 2025, 2:19:13 PMJun 25

to idtracker.ai users group

Hi, I'm trying to improve my video analysis process, and was wondering if there were recommendation for this?

My videos (90 15-minute videos, each segmented into ~10 random 10s .avi clips, or ~900 clips!) are of ~100-200um larvae swimming against a black background, and depending on the light conditions they are picked up at a size on the videos of 5-300pixels. These larvae are difficult to grow, small, and swimming in 3D space, and it's not possible for me to take additional videos.

So far, I have been successful in using the exclusive_rios with bounding boxes in the segmenter app to get clean tracks of the larvae. I use this after watching the clip at 36X speed, to determine that the animal is generally present throughout the entire clip. The larvae tend to "shimmer" and the lighting conditions differ with each clip, so the "background difference threshold" and "blob area threshold" is done very manually for each clip/video within a small range of values (5-30 background difference threshold, 5-8 lower limit and 60-300 upper limit blob area threshold).

I recently discovered that in the new idtracker.ai 6.0.2 I'm able to simply run a file with a low contrast value and get an output of 12 tracks, with 2 of the 12 being the good ones, which is great because I think I will be able to filter the bad tracks out with code rather than manually.

Is there a way to run the segmenter app with a series of settings / bash / toml files and select the best parameters for the video to best automate it? Or, I can run with the lowest common denominator settings (i.e. 5 background difference threshold, 5 lower limit and 300 upper limit blob area threshold) and select only the good runs.

Are there any recommendation for optimizing this process? I'm attaching an image of the bad track and good track from the same run.

Thanks for your time,

-Molly

“Screen_Shot” 2025-06-25 at 1.07.09 PM.png

“Screen_Shot” 2025-06-25 at 1.06.17 PM.png

Molly Roberts

unread,

Jun 25, 2025, 2:47:00 PMJun 25

to idtracker.ai users group

Just to clarify -- (and I'm only asking because version 6.0.3 is working so well!)

What if I were to run an in-line version of the segmentation app with a series of starting values for "background difference threshold" to get say ~10-20 animals, and then remove "bad trajectories" based on jumping behavior. Are there any functions built in that would do this for me?

The idtracker_validate gets rid of bad points, not bad trajectories (as far as I know), but I could do this in R / Python.

Is there a way to use a series of starting inputs and have the number of animals as outputs to avoid having to specify this manually? I'm considering writing something in bash...

Thanks for considering,

-Molly

idtracker.ai

unread,

Jun 26, 2025, 5:48:22 AMJun 26

to molly.a...@gmail.com, idtracker.ai users group

Hi Molly,

Setting a higher number of animals than the real one to then filter out bad trajectories is not the standard but a very interesting way of using idtracker.ai. I never tried it so I don't know what results out of this and for sure the Validator won't help you on that cause it assumes all trajectories correspond to different animals. But please tell me how well this workflow works for you and your workarounds, I'm interested :)

You can automate running idtracker.ai with bash scripts (easiest way if you know bash) or even Python scripts. For that you can run some loops where you set your parameters either into a toml file to load it after using "--load", or directly in the command line call like "idtrackerai --load basic_parameters.toml --number_of_animals 634 --intensity_ths 20 255 --name test_with_634_animals ... --track". Use the flag "--track" to disable the segmentation GUI and start the tracking directly, and "--name" not to overwrite the same session folder.

After this batch processing, we don't have available tools to postprocess the way you need...

Apart from this, I don't get why the tracking is so bad in your screenshots. If possible could you share with me (in...@idtracker.ai) one of your clips to try it by myself?

Jordi

--
You received this message because you are subscribed to the Google Groups "idtracker.ai users group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to idtrackerai_us...@googlegroups.com.
To view this discussion, visit https://groups.google.com/d/msgid/idtrackerai_users/e5d106c2-1099-454d-83ee-948248c117dcn%40googlegroups.com.

Molly Roberts

unread,

Jul 2, 2025, 4:20:16 PMJul 2

to idtracker.ai users group

Thanks for the advice!

I decided to work on optimizing the toml file automatically using 1 bash and 2 python files, however am having a new problem, that my new automatically generated ROIs are displaying correctly in idtrackerai and idtrackerai_validate, but exclusive_rois = True is no longer working, even though I do get Global Fragments. I have been using v6.0.3. Any recommendations? I've tried a few ways of getting exclusive_rois to work (through the CLI or through the toml file or through the browser OR both!), but there are tracks that are not constrained to the ROIs. Any recommendations?

Loading parameters from /gpfs/home/emroberts/toml_files/scallop_T2/test_automation/clip_1_00:11:20.toml: py_utils.py:83

video_paths=['/gpfs/scratch/emroberts/raw_videos/scallop/10s_clips/Time_2/3_14_7.3/Tank_2_14_7.3_

highlight_cam1/clip_1_00:11:20.avi']

intensity_ths=[10, 255]

area_ths=[5, 200]

tracking_intervals=None

number_of_animals=3

use_bkg=True

check_segmentation=False

track_wo_identities=True

roi_list=['+ Polygon [[57.921, 490.235], [57.921, 540.685], [82.629, 540.685], [82.629,

490.235]]',

'+ Polygon [[744.460, 4.333], [744.460, 59.833], [768.042, 59.833], [768.042,

4.333]]',

'+ Polygon [[483.430, 240.950], [483.430, 306.367], [506.555, 306.367], [506.555,

240.950]]',

'+ Polygon [[1001.741, 707.009], [1001.741, 756.097], [1026.946, 756.097],

[1026.946, 707.009]]',

'+ Polygon [[797.079, 907.891], [797.079, 959.296], [827.605, 959.296], [827.605,

907.891]]',

'+ Polygon [[451.515, 40.012], [451.515, 103.244], [487.444, 103.244], [487.444,

40.012]]',

'+ Polygon [[891.250, -9.399], [891.250, 30.506], [913.849, 30.506], [913.849,

-9.399]]',

'- Polygon [[1100.4, 8.7], [1115.0, 1074.7], [1119.3, 1101.9], [1932.7, 1101.9],

[1932.7, -18.4], [1098.4, -16.3]]']

background_subtraction_stat='min'

Terminal arguments: __main__.py:47

name='clip_1_00:11:20_run_thresh_10_roi'

video_paths=['/gpfs/scratch/emroberts/raw_videos/scallop/10s_clips/Time_2/3_14_7.3/Tank_2_14_7

.3_highlight_cam1/clip_1_00:11:20.avi']

intensity_ths=[10.0, 255.0]

area_ths=[5.0, 200.0]

number_of_animals=8

use_bkg=True

exclusive_rois=True

track_wo_identities=True

background_subtraction_stat='min'

number_of_frames_for_background=150

14:56:13 WARNING A new release of idtracker.ai is available: 6.0.3 -> 6.0.5 telemetry.py:117

To update, run: "python -m pip install --upgrade idtrackerai"

14:56:14 Using PyTorch 2.7.1+cu128 device.py:13

Using Cuda backend with "NVIDIA A100-PCIE-40GB" device.py:26

14:56:16 START Tracking session run.py:26

Initializing Session session.py:156

Setting video paths to: session.py:168

/gpfs/scratch/emroberts/raw_videos/scallop/10s_clips/Time_2/3_14_7.3/Tank_2_14_7.3_highlight_cam1/clip_1_00:1

1:20.avi

The session has 600 frames (2 episodes) session.py:222

Episode 0, frames (0 -> 300) of /clip_1_00:11:20.avi session.py:229

Episode 1, frames (300 -> 600) of /clip_1_00:11:20.avi session.py:229

Directory session.py:261

/gpfs/scratch/emroberts/raw_videos/scallop/10s_clips/Time_2/3_14_7.3/Tank_2_14_7.3_highlight_cam1/session_clip_1_

00:11:20_run_thresh_10_roi already exists

Directory session.py:262

/gpfs/scratch/emroberts/raw_videos/scallop/10s_clips/Time_2/3_14_7.3/Tank_2_14_7.3_highlight_cam1/session_clip_1_

00:11:20_run_thresh_10_roi/preprocessing already exists

ROI mask saved at session.py:590

/gpfs/scratch/emroberts/raw_videos/scallop/10s_clips/Time_2/3_14_7.3/Tank_2_14_7.3_highlight_cam1/session_clip_1_

00:11:20_run_thresh_10_roi/preprocessing/ROI_mask.png

Number of parallel jobs: 4 session.py:284

Saving Session object in run.py:112

/gpfs/scratch/emroberts/raw_videos/scallop/10s_clips/Time_2/3_14_7.3/Tank_2_14_7.3_highlight_cam1/session_clip_1_00:1

1:20_run_thresh_10_roi/session.json

START Animal detection py_utils.py:196

Directory animals_detection.py:17

/gpfs/scratch/emroberts/raw_videos/scallop/10s_clips/Time_2/3_14_7.3/Tank_2_14_7.3_highlight_cam1/sessio

n_clip_1_00:11:20_run_thresh_10_roi/bounding_box_images has been created

Using previously computed background model from GUI animals_detection.py:29

Segmenting video, 2 episodes in 4 parallel jobs segmentation.py:199

Saving bounding box images in segmentation.py:202

/gpfs/scratch/emroberts/raw_videos/scallop/10s_clips/Time_2/3_14_7.3/Tank_2_14_7.3_highlight_cam1/session_cl

ip_1_00:11:20_run_thresh_10_roi/bounding_box_images

14:56:20 Segmenting video (2 iterations). It took 0:00:03 segmentation.py:228

Generating ListOfBlobs object list_of_blobs.py:48

2256 detected blobs in total, an average of 3.8 blobs per frame animals_detection.py:57

WARNING There are no frames where the number of blobs is equal to the number of animals stated by the animals_detection.py:86

user 8. Depending on the video this is inevitable and idtracker.ai can handle it. But, if possible,

check the segmentation parameters to make sure your animals are visible and properly segmented.

There are 0 frames with more blobs than animals animals_detection.py:101

FINISH Animal detection, it took 0:00:04 py_utils.py:201

Saving Session object in run.py:112

/gpfs/scratch/emroberts/raw_videos/scallop/10s_clips/Time_2/3_14_7.3/Tank_2_14_7.3_highlight_cam1/session_clip_1_00:1

1:20_run_thresh_10_roi/session.json

Saving ListOfBlobs at run.py:114

/gpfs/scratch/emroberts/raw_videos/scallop/10s_clips/Time_2/3_14_7.3/Tank_2_14_7.3_highlight_cam1/session_clip_1_00:1

1:20_run_thresh_10_roi/preprocessing/list_of_blobs.pickle

Writing list_of_blobs.pickle (326.6 kB). It took 0:00:00 list_of_blobs.py:100

START Crossing detection py_utils.py:196

Computing body lengths (600 iterations). It took 0:00:00 model_area.py:75

Computing body lengths (2256 iterations). It took 0:00:00 model_area.py:83

The median body length is 7.8 pixels model_area.py:89

The automatic identification image size is 6 session.py:492

No resolution reduction required session.py:505

Identification image size set to [6, 6, 1], resolution reduction factor set to 1 session.py:540

Directory __init__.py:27

/gpfs/scratch/emroberts/raw_videos/scallop/10s_clips/Time_2/3_14_7.3/Tank_2_14_7.3_highlight_cam1/session_clip_1_

00:11:20_run_thresh_10_roi/identification_images has been emptied

14:56:21 Setting images for identification (2 iterations). It took 0:00:00 list_of_blobs.py:226

Connecting blobs (600 iterations). It took 0:00:00 list_of_blobs.py:161

Classifying Blobs as individuals or crossings depending on their area and the number of blobs in the crossing_detector.py:31

frame

Initializing ModelArea for individual/crossing blob initial classification model_area.py:40

Model area computed with 2256 blobs: mean=18.3, median=18.0, and std=10.0 pixels model_area.py:59

2256 blobs seem like individuals, 0 seem like crossings crossing_detector.py:46

Get list of blobs for training, validation and eval crossings_dataset.py:18

First individual/crossing assignment (600 iterations). It took 0:00:00 crossings_dataset.py:22

0 sure individuals and 0 sure crossings crossings_dataset.py:46

0 individual and 0 crossing blobs for training crossings_dataset.py:83

0 individual and 0 crossing blobs for validation

2256 unknown blobs crossing_detector.py:98

There are not enough crossings to train the crossing detector crossing_detector.py:104

FINISH Crossing detection, it took 0:00:01 py_utils.py:201

Saving Session object in run.py:112

/gpfs/scratch/emroberts/raw_videos/scallop/10s_clips/Time_2/3_14_7.3/Tank_2_14_7.3_highlight_cam1/session_clip_1_00:1

1:20_run_thresh_10_roi/session.json

Saving ListOfBlobs at run.py:114

/gpfs/scratch/emroberts/raw_videos/scallop/10s_clips/Time_2/3_14_7.3/Tank_2_14_7.3_highlight_cam1/session_clip_1_00:1

1:20_run_thresh_10_roi/preprocessing/list_of_blobs.pickle

Writing list_of_blobs.pickle (376.4 kB). It took 0:00:00 list_of_blobs.py:100

START Fragmentation py_utils.py:196

Finding blob's exclusive ROI (2256 iterations). It took 0:00:00 fragmentation.py:85

Fragmenting blobs (600 iterations). It took 0:00:00 fragmentation.py:59

Creating list of fragments list_of_fragments.py:191

Connecting coexisting fragments (264 iterations). It took 0:00:00 list_of_fragments.py:432

264 Fragments in total, 264 individuals and 0 crossings fragmentation.py:30

Total number of Global Fragments: 0 list_of_global_fragments.py:78

Of which 0 are long enough to be accumulated list_of_global_fragments.py:82

FINISH Fragmentation, it took 0:00:00 py_utils.py:201

Saving Session object in run.py:112

/gpfs/scratch/emroberts/raw_videos/scallop/10s_clips/Time_2/3_14_7.3/Tank_2_14_7.3_highlight_cam1/session_clip_1_00:1

1:20_run_thresh_10_roi/session.json

Saving ListOfBlobs at run.py:114

/gpfs/scratch/emroberts/raw_videos/scallop/10s_clips/Time_2/3_14_7.3/Tank_2_14_7.3_highlight_cam1/session_clip_1_00:1

1:20_run_thresh_10_roi/preprocessing/list_of_blobs.pickle

Writing list_of_blobs.pickle (385.5 kB). It took 0:00:00 list_of_blobs.py:100

Saving ListOfFragments as run.py:116

/gpfs/scratch/emroberts/raw_videos/scallop/10s_clips/Time_2/3_14_7.3/Tank_2_14_7.3_highlight_cam1/session_clip_1_00:1

1:20_run_thresh_10_roi/preprocessing/list_of_fragments.json

14:56:22 Writing list_of_fragments.json (123.4 kB). It took 0:00:00 list_of_fragments.py:242

Saving ListOfGlobalFragments at run.py:118

/gpfs/scratch/emroberts/raw_videos/scallop/10s_clips/Time_2/3_14_7.3/Tank_2_14_7.3_highlight_cam1/session_clip_1_00:1

1:20_run_thresh_10_roi/preprocessing/list_of_global_fragments.json

START Tracking py_utils.py:196

Tracking without identities __init__.py:59

Assigning random identities (600 iterations). It took 0:00:00 __init__.py:64

FINISH Tracking, it took 0:00:00 py_utils.py:201

Saving Session object in run.py:112

/gpfs/scratch/emroberts/raw_videos/scallop/10s_clips/Time_2/3_14_7.3/Tank_2_14_7.3_highlight_cam1/session_clip_1_00:1

1:20_run_thresh_10_roi/session.json

Saving ListOfBlobs at run.py:114

/gpfs/scratch/emroberts/raw_videos/scallop/10s_clips/Time_2/3_14_7.3/Tank_2_14_7.3_highlight_cam1/session_clip_1_00:1

1:20_run_thresh_10_roi/preprocessing/list_of_blobs.pickle

Writing list_of_blobs.pickle (394.5 kB). It took 0:00:00 list_of_blobs.py:100

Saving ListOfFragments as run.py:116

/gpfs/scratch/emroberts/raw_videos/scallop/10s_clips/Time_2/3_14_7.3/Tank_2_14_7.3_highlight_cam1/session_clip_1_00:1

1:20_run_thresh_10_roi/preprocessing/list_of_fragments.json

Writing list_of_fragments.json (123.4 kB). It took 0:00:00 list_of_fragments.py:242

Saving ListOfGlobalFragments at run.py:118

/gpfs/scratch/emroberts/raw_videos/scallop/10s_clips/Time_2/3_14_7.3/Tank_2_14_7.3_highlight_cam1/session_clip_1_00:1

1:20_run_thresh_10_roi/preprocessing/list_of_global_fragments.json

START Trajectories creation py_utils.py:196

Directory trajectories_creation.py:80

/gpfs/scratch/emroberts/raw_videos/scallop/10s_clips/Time_2/3_14_7.3/Tank_2_14_7.3_highlight_cam1/se

ssion_clip_1_00:11:20_run_thresh_10_roi/trajectories has been emptied

Producing trajectories (600 iterations). It took 0:00:00 trajectories_creation.py:129

Saving trajectories in trajectories_io.py:87

/gpfs/scratch/emroberts/raw_videos/scallop/10s_clips/Time_2/3_14_7.3/Tank_2_14_7.3_highlight_cam1/session_

clip_1_00:11:20_run_thresh_10_roi/trajectories/trajectories.h5

Creating converter from 5 to 3 attrs.py:205

Saving trajectories in trajectories_io.py:40

/gpfs/scratch/emroberts/raw_videos/scallop/10s_clips/Time_2/3_14_7.3/Tank_2_14_7.3_highlight_cam1/session_

clip_1_00:11:20_run_thresh_10_roi/trajectories/trajectories.npy

Saving trajectories in trajectories_io.py:53

/gpfs/scratch/emroberts/raw_videos/scallop/10s_clips/Time_2/3_14_7.3/Tank_2_14_7.3_highlight_cam1/session_

clip_1_00:11:20_run_thresh_10_roi/trajectories/trajectories_csv

Directory trajectories_io.py:54

/gpfs/scratch/emroberts/raw_videos/scallop/10s_clips/Time_2/3_14_7.3/Tank_2_14_7.3_highlight_cam1/session_

clip_1_00:11:20_run_thresh_10_roi/trajectories/trajectories_csv has been created

FINISH Trajectories creation, it took 0:00:00 py_utils.py:201

FINISH Tracking session, it took 0:00:05 run.py:67

Saving Session object in run.py:112

/gpfs/scratch/emroberts/raw_videos/scallop/10s_clips/Time_2/3_14_7.3/Tank_2_14_7.3_highlight_cam1/session_clip_1_00:1

1:20_run_thresh_10_roi/session.json

Saving ListOfBlobs at run.py:114

/gpfs/scratch/emroberts/raw_videos/scallop/10s_clips/Time_2/3_14_7.3/Tank_2_14_7.3_highlight_cam1/session_clip_1_00:1

1:20_run_thresh_10_roi/preprocessing/list_of_blobs.pickle

Writing list_of_blobs.pickle (394.5 kB). It took 0:00:00 list_of_blobs.py:100

Saving ListOfFragments as run.py:116

/gpfs/scratch/emroberts/raw_videos/scallop/10s_clips/Time_2/3_14_7.3/Tank_2_14_7.3_highlight_cam1/session_clip_1_00:1

1:20_run_thresh_10_roi/preprocessing/list_of_fragments.json

Writing list_of_fragments.json (123.4 kB). It took 0:00:00 list_of_fragments.py:242

Saving ListOfGlobalFragments at run.py:118

/gpfs/scratch/emroberts/raw_videos/scallop/10s_clips/Time_2/3_14_7.3/Tank_2_14_7.3_highlight_cam1/session_clip_1_00:1

1:20_run_thresh_10_roi/preprocessing/list_of_global_fragments.json

Tracked without identities, no estimated accuracy available. run.py:71

Data policy: 'idmatcher.ai' session.py:951

Directory session.py:970

/gpfs/scratch/emroberts/raw_videos/scallop/10s_clips/Time_2/3_14_7.3/Tank_2_14_7.3_highlight_cam1/session_clip_1_

00:11:20_run_thresh_10_roi/bounding_box_images has been removed

Directory session.py:971

/gpfs/scratch/emroberts/raw_videos/scallop/10s_clips/Time_2/3_14_7.3/Tank_2_14_7.3_highlight_cam1/session_clip_1_

00:11:20_run_thresh_10_roi/crossings_detector not found, can't remove

Compressing identification images (2 iterations). It took 0:00:00 session.py:987

Success run.py:81

Log file copied to run.py:106

/gpfs/scratch/emroberts/raw_videos/scallop/10s_clips/Time_2/3_14_7.3/Tank_2_14_7.3_highlight_cam1/session_clip_1_00:1

1:20_run_thresh_10_roi/idtrackerai.log

idtracker.ai

unread,

Jul 3, 2025, 12:38:03 PMJul 3

to molly.a...@gmail.com, idtracker.ai users group

Hi,

Exclusive ROIs were not intended to work with "tracking without identities". "Tracking without identities" was designed as an extra option that bypasses most of the idtrackerai processes and error corrections.

Since you asked, I just released version 6.0.6 including exclusive ROIs to work with "tracking without identities". Give it a try!

Jordi

To view this discussion, visit https://groups.google.com/d/msgid/idtrackerai_users/b94ad82a-8909-41d6-8c84-1989827f3e8cn%40googlegroups.com.

Molly Roberts

unread,

Jul 3, 2025, 5:43:21 PMJul 3

to idtracker.ai users group

Thanks again for your quick response.

I had an an issue upgrading from 6.0.3 to 6.0.6 because I am using PyQt5 (I'm on a linux machine and can't do the sudo fix for using PyQt6). We made 2 changes to the QT Icon call (see attached image below) and then the upgrade to 6.0.6 seems to have worked. I'm attaching my history below.

Thanks again! I'm looking forwared to trying this out!

-Molly

Here is my history...

1155 2025-07-03 16:38:34 conda create -n idtrackerai_py12 python=3.12.0 -y

1156 2025-07-03 16:39:42 conda activate idtrackerai_py12

1157 2025-07-03 16:39:58 whereis python

1158 2025-07-03 16:40:57 conda install pip

1159 2025-07-03 16:41:05 whereis pip

1160 2025-07-03 16:41:24 /gpfs/home/emroberts/.conda/envs/idtrackerai_py12/bin/pip install idtrackerai

1161 2025-07-03 16:41:42 /gpfs/home/emroberts/.conda/envs/idtrackerai_py12/bin/pip install PyQt6

1162 2025-07-03 16:41:50 idtrackerai_validate

1163 2025-07-03 16:42:25 conda list

1164 2025-07-03 16:42:58 conda install PyQt5

1165 2025-07-03 16:43:05 conda install pyqt5

1166 2025-07-03 16:43:22 /gpfs/home/emroberts/.conda/envs/idtrackerai_py12/bin/pip uninstall PyQt6

1167 2025-07-03 16:43:41 /gpfs/home/emroberts/.conda/envs/idtrackerai_py12/bin/pip install PyQt5

1168 2025-07-03 16:43:48 idtrackerai_validate

1169 2025-07-03 16:49:42 whereis pip

1170 2025-07-03 16:51:58 ls /gpfs/home/emroberts/.local/lib/python3.12/site-packages/idtrackerai/

1171 2025-07-03 16:52:22 pip list

1172 2025-07-03 16:52:36 /gpfs/home/emroberts/.conda/envs/idtrackerai_py12/bin/pip list

1173 2025-07-03 16:52:59 cd /gpfs/home/emroberts/.conda/envs/idtrackerai_py12

1174 2025-07-03 16:53:00 ls

1175 2025-07-03 16:53:03 cd lib/

1176 2025-07-03 16:53:03 ls

1177 2025-07-03 16:53:16 cd ..

1178 2025-07-03 16:53:17 ls

1179 2025-07-03 16:53:24 grep -R 'idtrack' ./

1180 2025-07-03 16:54:43 nano lib/python3.12/site-packages/idtrackerai/GUI_tools/widgets_utils/video_player.py

1181 2025-07-03 16:57:08 cd

1182 2025-07-03 16:57:14 idtrackerai_validate

1183 2025-07-03 16:59:13 nano /gpfs/home/emroberts/.conda/envs/idtrackerai_py12/lib/python3.12/site-packages/idtrackerai/GUI_tools/widgets_utils/video_player.py

1184 2025-07-03 16:59:40 idtrackerai_validate

1185 2025-07-03 17:01:04 nano /gpfs/home/emroberts/.conda/envs/idtrackerai_py12/lib/python3.12/site-packages/idtrackerai/GUI_tools/widgets_utils/video_player.py

1186 2025-07-03 17:02:21 idtrackerai_validate

“Screen_Shot” 2025-07-03 at 5.04.57 PM.png

idtracker.ai

unread,

Jul 4, 2025, 10:44:19 AMJul 4

to molly.a...@gmail.com, idtracker.ai users group

Oh no! So sorry, I didn't double check the compatibility with PyQt5. I released yet another version (6.0.7) with this hotfix. Hoping to stop releasing versions every week soon 😅.

Jordi

To view this discussion, visit https://groups.google.com/d/msgid/idtrackerai_users/f2280616-ba8d-4009-9b85-65edebdf077an%40googlegroups.com.

Molly Roberts

unread,

Jul 9, 2025, 4:26:35 PMJul 9

to idtracker.ai users group

Thanks! It's working! Some time soon I'll post the bash code I am using to automatically generate the ROI "arenas!"

Reply all

Reply to author

Forward