Comments (14)
Dear @pordeli ,
No. You need the confidence scores to evaluate your detections.
Best regards,
Rafael
from object-detection-metrics.
Thanks for your reply. As long as I do not have the scores do you know a kind of code that I can use to get precision and recall and the curve for my 100 frames of the video?
from object-detection-metrics.
hello ,
I try to evaluate the model with my own dataset .Can you tell me how to get the box coordinates detected in this format <class_name confidence left top right bottom>
Thanks
from object-detection-metrics.
@pordeli ,
Due to the behavior of the metric we apply you will need the confidences. I do not know any other code that you could use.
from object-detection-metrics.
@Ibmaria ,
For the detections you need to apply an object detector, such as YOLO, Faster R-CNN, Fast R-CNN, etc. They will provide the detection bounding boxes, classes and confidences.
from object-detection-metrics.
For the model i use ssd mobilenet ,
for evaluation you said that to create 2 folders for ground truth and detection .How did you create detection file in the format class_name, confidence left top right bottom .I can not save them in txt format .How to save them like ground truth.Thanks for advance
from object-detection-metrics.
@Ibmaria ,
I never worked with mobilenet, so I dont know its output format. But you should get its output and convert it to the required format.
from object-detection-metrics.
@rafaelpadilla
my question is that how to convert it to the required format with the same name 00001.txt
from object-detection-metrics.
Dear @Ibmaria ,
You will need to create a script for that. It will take your detections to convert from the format mobilenet provides to the format and name required.
Could you show me the detections format given by mobilenet?
from object-detection-metrics.
@rafaelpadilla
can you show me rather your script.I will try to understand and adapt it to mine.Or your email i will send ssen you my code .Thanks in advance
from object-detection-metrics.
@rafaelpadilla
Finally i did but I will be delighted to see yours .Thanks
from object-detection-metrics.
@Ibmaria ,
I do not have a script for that. If you provide the output format of the mobilenet, I might be able to help you.
from object-detection-metrics.
@Ibmaria ,
Was your problem solved?
from object-detection-metrics.
@rafaelpadilla
yes i did .Thanks
from object-detection-metrics.
Related Issues (20)
- need to cast classId to str when saving PlotPrecisionRecallCurve HOT 2
- Question about the argument -imgsize HOT 1
- Confidence threshold HOT 8
- NMS ?? HOT 3
- Possible bug in BoundingBox.py? HOT 1
- How to calculate mAP@[0.5:0.99]? HOT 2
- red boxes represent detections with prediction label of that class or any detections whenever it prediction is that class ? HOT 2
- incorrect equation HOT 4
- ggggg
- Check performance of the trained model HOT 2
- The way to set which object is TP when more than one detection overlapping a ground truth seems to be wrong HOT 2
- How can i get TP,TN,FP,FN from it? HOT 1
- [question] support for 3d volumes? HOT 1
- Difference implementations between this repo and the faster_rcnn ones HOT 1
- What are the dimensions of precision represent? HOT 1
- Got `AP=0.00%` when running `pascalvoc.py` with samples
- image3 G iou ? HOT 2
- getting threshold values of confidence score which recall/precision calculated HOT 3
- I think a bug resulting in incorrect calculation for Average Precision HOT 1
- Definition of xywh is <left> <top> <width> <height> or <center-x> <center-y> <width> <height>
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 object-detection-metrics.