Comments (9)
This error is usually caused by unmatched predict_pid & i & real_tracks range . More informations of the variables are needed.
- which target dataset is used?
- which variable is the error index? i or predict_pid? what's the shape of real_tracks?
- have you correctly prepared track info file(corresponding to answer_path) for your target dataset? this file is parsed before computing spatial temporal pattern. Parsing rules are different for different dataset.
Furthermore, your code seems to be different from the published version according to different line number (93 vs 78)of the crash line. More modification details should be provided.
from tfusion.
Thank you for quick reply.
The target dataset is grid_cross0.
i and predict_pid are 0, 344.
real_tracks shape is (250, 4)
I know id index out of list.
I thought it's the dataset problem.
Could give me your experiment dataset from baiduyunpan or others?
For the track info file, I didn't prepare it, Could you give a sample?
For the different line problem, I write the 'print()' codes, no damage to the publish one.
from tfusion.
predict_pid is created from renew_pid, meaning the file in such path: TrackViz/data/market_grid_cross0-train/renew_pid.log. This file is generated by the visual classifier, you need to make the visual classifier work on the right target dataset's training set. How many images are there in your grid_cross0/train directory?
from tfusion.
1275 images in grid_cross0/train directory, it should be same size and image id as your data file?
from tfusion.
It's different. GRID claims a 10-time-2-fold cross-validation, so we divide the dataset as they suggest. Details of train-probe-gallery division is listed here: https://github.com/ahangchen/TFusion/tree/master/TrackViz/data/grid
from tfusion.
Under your guidance, the issue before has been solved.
However, bring out a new issue.
Output is that:
/home/msc/Github/person_reid/TFusion/TrackViz/data/market_grid-cv0-test/renew_pid.log
[0.0, 0.008, 0.016, 0.016, 0.032]
left image ready
predict images ready
deltas collected
deltas sorted
deltas saved
left image ready
predict images ready
deltas collected
deltas sorted
deltas saved
left image ready
predict images ready
deltas collected
deltas sorted
deltas saved
fusion on training dataset
img score ready
above person score ready
0.02941583853920899
fusion on test dataset
copy train track distribute pickle done
probe and gallery tracks ready
read vision scores and pids ready
load track deltas ready
load rand deltas ready
load diff deltas ready
Traceback (most recent call last):
File "ctrl/transfer.py", line 133, in
dataset_fusion_transfer()
File "ctrl/transfer.py", line 126, in dataset_fusion_transfer
fusion_transfer(source, 'grid-cv%d' % i)
File "ctrl/transfer.py", line 99, in fusion_transfer
fusion_test_rank_pids_path, fusion_test_rank_scores_path = st_fusion(source, target)
File "ctrl/transfer.py", line 48, in st_fusion
init_strict_img_st_fusion()
File "/home/msc/Github/person_reid/TFusion/TrackViz/ctrl/img_st_fusion.py", line 74, in init_strict_img_st_fusion
iter_strict_img_st_fusion(on_test=True)
File "/home/msc/Github/person_reid/TFusion/TrackViz/ctrl/img_st_fusion.py", line 115, in iter_strict_img_st_fusion
test_fusion(fusion_param)
File "/home/msc/Github/person_reid/TFusion/TrackViz/ctrl/img_st_fusion.py", line 29, in test_fusion
fusion_st_gallery_ranker(fusion_param)
File "./train/st_filter.py", line 303, in fusion_st_gallery_ranker
rand_track_scores = gallery_track_scores(query_tracks, gallery_tracks, rand_delta_s, fusion_param)
File "./train/st_filter.py", line 224, in gallery_track_scores
time2 = gallery_tracks[pid4probe][2]
IndexError: list index out of range
from tfusion.
I think this is caused by similar issues on gallery images. Please make sure that your image number in cross0/gallery is the same as the division in my last comment.
from tfusion.
However, all cross folders are same as the description files in your repo.
But, every line in the gallery.txt file is same which txt file? grid-cv%d-gallery.txt or grid-cv%d-test.txt?
from tfusion.
All cross folders are different.You would know GRID Gallery images share 775 unlabeled images and 125 labeled images in gallery for each cross if you have read GRID dataset's instruction. I think you only glance at the gallery file and only notice the shared 775 unlabeled images
while ignore 125 different labeled images
in each cross.
I'm sorry that grid-cv%d-gallery.txt is exactly the same with grid-cv%d-test.txt, but we only use grid-cv%d-gallery.txt so it doesn't matter.
from tfusion.
Related Issues (20)
- 中文讨论请到这里 HOT 70
- 运行pair_train.py文件失败 HOT 1
- question HOT 34
- M1,M2,M3 HOT 11
- #新手学习,请多多指教# HOT 1
- market-grid视觉迁移 HOT 1
- .log 相关问题 HOT 1
- No such file or directory: 'data/market_grid-cv0-train/renew_pid.log' HOT 2
- .log HOT 1
- ImportError: No module named baseline.train
- 中
- transfer.py 运行有错,不太懂 HOT 1
- test生成文件不全 HOT 1
- Number of people HOT 1
- Market 1501 dataset link broken HOT 1
- About Dataset HOT 2
- Pre-trained models and real-time tracking
- 关于 rank1.rank5,rank10
- 关于“answer_path” HOT 2
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from tfusion.