I am CEO of my own company related to AI and Computer Vision.
I want to train my model on my own image dataset to detect certain subjects using MXNet or Gluon.
If you have experienced in CNN like TensorFlow or yolo, it is easy to understand.
[login to view URL]@manivannan_data/how-to-train-multiple-objects-in-yolov2-using-your-own-dataset-2b4fee898f17
For example, this tutorial shows how to train your own convolutional neural network object detection classifier for dog and cat, starting from scratch.
I want to implement yolo with MXNet or Gluoncv.
The dataset is not a problem. You can choose your own dataset or download the useful dataset from Google.
I don't need the trained model. I need the process of how to set up the dataset and train the model.
There are some references:
[login to view URL]
This shows how to prepare custom datasets for object detection in Gluon.
[login to view URL]
This shows how to train YOLOv3 on pascal VOC using GluonCV.
After training, I want to import and use the trained model in my c++ or java project.
Eventually, the milestones will be like this.
1. configure some steps for preparing dataset and training the model.
You have to teach me all stage using documentation or remoted program on my computer.
I have experience in ML, so you don't need to make much more details.
deadline: 3 days
2. develop a test program using c++ or java that is for test the trained model. Input is one image and output is the image of detected objects with rectangle coordinate.
deadline 3 days
3. integrate the model and test program into my main project.
I will train the model with my own dataset according to your teaching.
The accuracy is the key here.
I guess the accuracy is not enough only with the trained model, so need more extra steps such as image processing like adjust and threshold.
The detail of my total project will be provided to the right person who would finish above 2 milestones. The details would be private and can't be published.
deadline: 7 days
Well, the first milestone budget may be low for you if you are an expert, but I can not pay more because the training model is just the beginning of our whole project. If you can lead our project into a success, I am willing to pay you more as you want.
Start your bid with "GluonCV and MXnet".
Please read more carefully the job description, or I will not accept your proposal.
27 freelancere byder i gennemsnit $21/time for dette job
Hi, Sir/Ma'am I have worked on some Image processing projects which require implementation of CNN. I have used YOLOv3, ResNet50, InceptionV6, etc. Ping me for getting more info of my projects done in this feild.