Comments (8)
https://everitt257.github.io/post/2018/08/10/object_detection.html
from nfl_kaggle.
@gsvsumanth please add your ideas here.
from nfl_kaggle.
well, you are trying to maintain the higher accuracy and reduce complexity. Faster RCNN is using VGG as a backbone model and is a two-stage object detector i.e. RPN and VGG while YOLO(Darknet backbone) is a single shot detector utilizing anchor boxes. one possible combination is to try using Darknet in faster RCNN replacing VGG and vice versa for YOLO. the other possible option is to remove RPN and utilize anchor boxes similar to YOLO in faster RCNN.
But again, it depends on your problem. what kind of dataset you want to use. like YOLO results are extremely bad on the self-driving cars dataset.
from nfl_kaggle.
from nfl_kaggle.
https://towardsdatascience.com/faster-r-cnn-a-step-towards-real-time-object-detection-98c186732a69
from nfl_kaggle.
well, you are trying to maintain the higher accuracy and reduce complexity. Faster RCNN is using VGG as a backbone model and is a two-stage object detector i.e. RPN and VGG while YOLO(Darknet backbone) is a single shot detector utilizing anchor boxes. one possible combination is to try using Darknet in faster RCNN replacing VGG and vice versa for YOLO. the other possible option is to remove RPN and utilize anchor boxes similar to YOLO in faster RCNN.
But again, it depends on your problem. what kind of dataset you want to use. like YOLO results are extremely bad on the self-driving cars dataset.
This is for real time detection
from nfl_kaggle.
Ideas
In order to avoid the anomalies ( like detecting objects other than helmets in a image and also finding a person who is missing in a frame (x+1) and reappearance of that person in (x+10) ) in object detection in real time using yolo v5
starting with speed and acceleration of each player we will track each player movement
there is a High chance that our player might get colloid in such a case the speed of the player will change this drastic change will help us in finding out whether there is Collision or not.
From end zone or from side zone the player helmets will not be able view as there are situation were they are hidden beside the other player in such a situation we tackle this problem by taking the number of helmets present in that frame (X) compared with the average number of players present in the (X+5) and (X-5) if it is equal to then we can take out
we can take the count of number of helmets from starting of a frame if the count of the number of helmets for a particular frame has increased we can look into frames were these anomalies have occurred
from nfl_kaggle.
@gsvsumanth please close this issue if there's nothing more to add.
from nfl_kaggle.
Related Issues (10)
- :sparkles: **Welcome to GitHub Projects** :sparkles:
- Please move the model_selection.md document to Docs folder HOT 1
- Making a middleware to use the confidence threshold as a variable to train the data. HOT 1
- Evaluation of trained models (ref: https://towardsdatascience.com/various-ways-to-evaluate-a-machine-learning-models-performance-230449055f15) HOT 3
- Creating the image cropping to detect the helmets. Comparing multiple notebooks and research papers listed. HOT 2
- Checking detectron Model HOT 1
- Training and Testing
- detectron2 runtime issues, in kaggle environment. HOT 2
- Things we can try from solutions
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 nfl_kaggle.