Comments (14)
I am not sure I understand you question here. If you are talking about the external scores, they are from an action recognition model and not necessary for most of the datasets.
from actionformer_release.
I am not sure I understand you question here. If you are talking about the external scores, they are from an action recognition model and not necessary for most of the datasets.
I mean when you run on the test dataset?when I run on test dataset ,I got the value of ‘output[vid_idx]['segments'][:, 0]’ is [-1., -1., -1., ..., -1., -1., -1.],I didn't know why? could you give me some advices?thx!!!!!
from actionformer_release.
To run on your own test set with unlabeled videos, you will need to
- extract video features the same way as the training set
- create a json file following the AcitivtyNet format with each annotation as an empty list
- copy the configfile used for training, and modify the fields "json_file" / " feat_folder" to point to the new json file and video features
- if "external_score_file" was specified in the original config, remove the filed and set "multiclass_nms" to True (disable the use of external classification scores)
- modify the eval.py to save the results into your desired format.
from actionformer_release.
from actionformer_release.
The segment means the real action starting and ending time (in seconds).
You may want to refer to the https://github.com/happyharrycn/actionformer_release/blob/main/libs/modeling/meta_archs.py#L643 for more details about how we convert the raw model outputs into final predictions.
from actionformer_release.
The segment means the real action starting and ending time (in seconds).
You may want to refer to the https://github.com/happyharrycn/actionformer_release/blob/main/libs/modeling/meta_archs.py#L643 for more details about how we convert the raw model outputs into final predictions.
I mean to submit the results in this web,https://codalab.lisn.upsaclay.fr/,Nouns and verbs submitted separately?
from actionformer_release.
You may need to refer to the specfic submission format for details. I think they should be submitted separately.
from actionformer_release.
You may need to refer to the specfic submission format for details. I think they should be submitted separately.
I can't open this URL about the specfic submission formats,could you?
from actionformer_release.
I think you may need to email the competition organizer to report this issue.
from actionformer_release.
If you are looking for combining noun and verb predictions, you might want to check the discussion in issue #7. If you have questions for EPIC-Kitchens data format, please contact the competition organizer.
from actionformer_release.
If you are looking for combining noun and verb predictions, you might want to check the discussion in issue #7. If you have questions for EPIC-Kitchens data format, please contact the competition organizer.
I am so confused ,both nouns and verbs predict segment time, how do I combine them to generate actions?
from actionformer_release.
For this part, you can have various choices to fuse these results.
For example, you can get the noun/verb predictions for the same point, then take the noun or verb segment predictions as the final segment for this point, or you can simply take the average of these two segments. You may need to modify the code a little bit.
We may update this part shortly.
from actionformer_release.
thxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx ! amazing!
from actionformer_release.
Mark as resolved.
from actionformer_release.
Related Issues (20)
- Possible to get rid off regression head? HOT 4
- Use which checkpoint to report performance when evaluating HOT 1
- EMA model not working HOT 2
- anet_1.3_i3d.tar.gz HOT 3
- May I ask why there is warn: "No predictions of label '16' were provdied". HOT 5
- question about input i3D features HOT 5
- Assertion error
- Error "predicted offsets must be non-negative" HOT 1
- multigpu training HOT 2
- How to extract features of an action after transformer encoder in pyramid architecture? HOT 4
- No module named 'nms_1d_cpu HOT 2
- SlowFast feature extraction HOT 7
- What's the unit of the maximum input sequence (max_seq_len) HOT 7
- Visualization appendix D HOT 1
- Including SlowFast in LocPointTransformer HOT 7
- Why replace the predicted labels? HOT 2
- Can we run evaluate script on multiple GPUS? HOT 1
- Regarding Epic-Kitchen Combination of Noun and Verb HOT 2
- Inference on single video HOT 2
- Guidelines on setting up config for custom datasets? HOT 5
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 actionformer_release.