You will get an image-based similarity recommender given an uploaded image
Project details
You will get an image-based similarity recommender for your uploaded image. With 2 years of experience in applying computer vision for retail companies, I bring deep learning ready for your enterprise to contribute to their growth and innovation through the latest AI techniques.
What's included $500
These options are included with the project scope.
$500
- Delivery Time 45 days
- Number of Revisions 1
- Number of Model Variations 0
- Number of Scenarios 0
- Model Validation/Testing
- Source Code
About Antonio de Jesus
Data Scientist
Cuautitlan, Mexico - 6:32 am local time
The tools that I use are Python, R, SAS, SQL and Scala/Spark. Also machine learning algorithms (supervised and unsupervised) e.g. Xgboost, logistic regressions, k-means clustering, etc., deep learning e.g. recurrent neural networks, and time series algorithms e.g. Vector Autorregression, Arima, etc. Also Google Cloud Compute Engine (to run algorithms such as recurrent neural networks or Bayesian optimization in the cloud).
My activities include data analysis, data wrangling, text mining, webscraping, forecasts of time series, classification and regression tasks, deploying models as websites, among others.
Steps for completing your project
After purchasing the project, send requirements so Antonio de Jesus can start the project.
Delivery time starts when Antonio de Jesus receives requirements from you.
Antonio de Jesus works on your project following the steps below.
Revisions may occur after the delivery date.
Receive dataset of labeled images
Check how many images are provided and how well they are split into labels depending on the similarity needs of the client. Depending on the number of images and its labels, more images and/or time to label them properly might be required.
Build classifier using deep learning techniques
By applying a deep leaning algorithm (CNN) to the labeled dataset, a classifier is built in order to learn the features and differences between images.