A multi-scale Inception V3 Approach with efficient fine-tuning

A tissue slide annotated with cancerous tumor cells


Each year, more than 230,000 breast cancer patients in the U.S. hinge on whether cancer has metastasized. Metastasis detection is performed by pathologists reviewing large expanses of biological tissues. This process is labor-intensive and error-prone. Source

In this project, we aim to implement a multi-scale metastases classification model proposed in the paper Detecting Cancer Metastases on Gigapixel Pathology Images.


We use the CAMELYON-16 Multi Giga Pixel slides data. A set of 22 slides with significant tumors were sampled from the main dataset. Each slide has a corresponding mask that annotates the tumor region. We can access around 9 magnification levels.

Photo by Matthew Henry from Burst

A visual analysis of ProPublica’s NYPD Civilian Misconduct Dataset

Legislation 50-a, implemented in 1976, refuses disclosure of police records to the public to prevent criminal defense lawyers from using them in cross-examination. On June 12th, 2020, the New York state repealed that law. ProPublica, a non-profit organization, submitted a records request and has made this data available to the general public subject to terms and conditions.

We analyzed this data to capture the trends in police brutality across years in NYC based on ethnicity and the corresponding impact of such cases on Police Officers.

The Dataset

We have used the Civilian Complaints Against New York City Police Officers dataset made available…

Smarth Gupta

MS in Data Science student @Columbia University | BS in Economics from @IITKanpur | Data Scientist @Spiny AI

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store