Comments (4)
The first three lines are the result of my own training model:
100 100 34.18 28.70 20.20 13.54 8.17 3.40 1.45 35 70 31.39 26.72 19.16 12.83 7.28 3.66 1.70 35 75 31.38 26.49 19.76 13.25 7.76 3.84 1.81
yours:
1000 1000 68.19 62.43 53.21 44.78 34.10 23.13 13.44
Our training settings are all presented in train_options.py, and there is no training trick in our method, you just need to train our model from scratch. Did you change some hyper-parameters?
from fac-net.
The first three lines are the result of my own training model:
100 100 34.18 28.70 20.20 13.54 8.17 3.40 1.45 35 70 31.39 26.72 19.16 12.83 7.28 3.66 1.70 35 75 31.38 26.49 19.76 13.25 7.76 3.84 1.81
yours:
1000 1000 68.19 62.43 53.21 44.78 34.10 23.13 13.44Our training settings are all presented in train_options.py, and there is no training trick in our method, you just need to train our model from scratch. Did you change some hyper-parameters?
I didn't change any hyper-parameters. Is the dataset Thumos14reduced-I3D-JOINTFeatures?
from fac-net.
The first three lines are the result of my own training model:
100 100 34.18 28.70 20.20 13.54 8.17 3.40 1.45 35 70 31.39 26.72 19.16 12.83 7.28 3.66 1.70 35 75 31.38 26.49 19.76 13.25 7.76 3.84 1.81
yours:
1000 1000 68.19 62.43 53.21 44.78 34.10 23.13 13.44Our training settings are all presented in train_options.py, and there is no training trick in our method, you just need to train our model from scratch. Did you change some hyper-parameters?
I didn't change any hyper-parameters. Is the dataset Thumos14reduced-I3D-JOINTFeatures?
Yes, I noticed the last line's result is the same as ours, so I think your features and annotations are right. Maybe you can change the random seed and train again. On our server, for the flow stream, the mAP at IoU 0.1 exceeds 50% after 10 epochs. Hope this information could be helpful to you.
from fac-net.
The first three lines are the result of my own training model:
100 100 34.18 28.70 20.20 13.54 8.17 3.40 1.45 35 70 31.39 26.72 19.16 12.83 7.28 3.66 1.70 35 75 31.38 26.49 19.76 13.25 7.76 3.84 1.81
yours:
1000 1000 68.19 62.43 53.21 44.78 34.10 23.13 13.44Our training settings are all presented in train_options.py, and there is no training trick in our method, you just need to train our model from scratch. Did you change some hyper-parameters?
I didn't change any hyper-parameters. Is the dataset Thumos14reduced-I3D-JOINTFeatures?
Yes, I noticed the last line's result is the same as ours, so I think your features and annotations are right. Maybe you can change the random seed and train again. On our server, for the flow stream, the mAP at IoU 0.1 exceeds 50% after 10 epochs. Hope this information could be helpful to you.
Thanks for your advice. I didn't realize that random seed is so important for small model.
from fac-net.
Related Issues (8)
- What is used by an existing network with added modules HOT 3
- About of extracted features HOT 1
- 关于本文 HOT 1
- Visualization HOT 2
- 关于训练 HOT 2
- requirements.txt HOT 1
- ActivityNet V1.3数据集的缺失问题 HOT 4
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 fac-net.