Comments (2)
Hi,
You are right about the changes in intention and it has to be updated throughout. In our work, we simplified the problem to increase the number of trajectories while predicting the intention of pedestrians on part of the data that was exactly the same as the data used in human experiments, i.e. up to the same critical points of prediction. Predicting beyond the critical points used in the experiments can be quite noisy because the intention model is not trained on such data. One way of handling this can be perhaps modifying the intention score based on the observation of the action, e.g. in this case if the crossing finished, we set the intention score very low (cannot be certain to set to 0), Another thing to consider is that when a pedestrian finishes crossing, s/he will very unlikely interact with the vehicle again, therefore s/he is not as important to the vehicle again.
from pie.
Hi @aras62 Thank you so much for your response! You have mentioned modifying the intention score annotations based on the observation of the action. My understanding is that to modify the intention score annotations, we need to mannually look at each trajectory. Please let me know if there is a way to make it automatic.
from pie.
Related Issues (20)
- annotation HOT 1
- Calibration not working? HOT 5
- Mean and standard deviation of images in the dataset
- Mean and standard deviation of images in dataset HOT 2
- About Attributes : intention_prob
- Is there a detailed explanation of the vehicle annotations? HOT 2
- Some images are missing cannot run the model HOT 1
- Pie Dataset website not accessible HOT 2
- How to use the dataset to train a YOLO object detector? HOT 5
- What does the bounding box coordinates mean? HOT 5
- The Annotations cannot match images HOT 3
- Annotation 'designated' HOT 2
- Annotation 'signalized' HOT 2
- about video to images HOT 2
- Speed Annotation unit HOT 2
- There may be an error in the pedestrian boundingbox annotation of set03video0001 HOT 2
- The label does not correspond to the video picture HOT 2
- Road Boundaries HOT 2
- what does '1_1_1' represent in annotation? HOT 1
- Clarification regarding the label "crossing" in the Annotation Attributes 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 pie.