Comments (3)
Please ignore the extracted_fps.npy file.
For I3D, we extract the video frames at 25fps and then extract a single feature vector for every 16 frame input to the network. We do the same thing for both the RGB and FLOW stream.
For UNT, we extract the video frames at 10 fps and then extract features for every frame (for flow it is a bit different and the input are 5 neighboring frames for every timestamp we want to extract the features). Then we pool 4 such features (non-overlapping) to get a single feature vector, which is then fed as input to our network (i.e., how the 2.5 features come from 10/4=2.5)
from wtalc-pytorch.
Thank you for the detailed reply! That totally solved my confusion!
from wtalc-pytorch.
Hello, I'm a newer, could you tell me that how to change the fps from 30 to 25 for each video? Should I need drop the frames whose index%6==0 ?
from wtalc-pytorch.
Related Issues (20)
- The model results vary greatly HOT 1
- I believe labels_all.npy is wrong HOT 1
- About Qualitative Results
- difference between labels.npy and labels_all.npy HOT 1
- Both Classification and Localization Performance is higher in my experiment than your eccv18 paper in Thumos14 HOT 4
- Specific parameter setting when training on activityNet v1.2 HOT 5
- How can i extract the feature by myself? HOT 1
- RuntimeError: The size of tensor a (101) must match the size of tensor b (20) at non-singleton dimension 1
- Visualization
- FPS of ActivityNet?
- Could you provide the pre-trained encoder?
- Regarding feature extraction HOT 2
- Can you share the UntrimmedNet feature for ActivityNet1.2
- regarding result
- Specific composition of features HOT 1
- calculate features HOT 13
- features download HOT 4
- training on ActivityNet v1.2 HOT 3
- overfitting HOT 1
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 wtalc-pytorch.