So in case you were wondering what Vidrovr is ...

On today's Web, video dominates online traffic (>80%) and multimedia has become standard in content creation and consumption. Not only is web traffic dominated by video, but online and mobile video advertising is the most rapidly growing and profitable advertising real estate available to media corporations and advertisers, and will only continue to grow into the future. To take advantage of this important avenue for monetization media corporations need systems that can index, search, and recommend their video content in a cost-effective, automatic, and accurate manner. Vidrovr provides those solutions. Vidrovr was founded by two PhD Students in the Digital Video and Multimedia Lab at Columbia University, Joseph Ellis and Daniel Morozoff. The Vidrovr team is advised by Prof. Shih-Fu Chang, of the Columbia School of Engineering and Applied Science.

Vidrovr has vast experience in building distributed systems that can process and learn from vast amounts of information in real time, in particular designing algorithms that utilize multiple facets of multimedia streams concurrently. They have published and patented foundational research in machine learning, computer vision, multimodal information processing, and multimedia. Over the past three years they have developed an award-winning and patented system for processing news videos and social media, called News Rover, and the technologies developed for this system are directly leveraged in Vidrovr products. Vidrovr was selected to participate in the NYC Media Lab’s start-up incubator, The Combine, and through this program completed over 110 customer interviews with media executives, video editors, producers, video management professionals, and more, to evaluate the pain-points that exist for large corporations in leveraging their video content. Vidrovr addresses 3 key market needs: 1. Domain and customer specific automatic metadata generation for videos, 2. Video Content Management solutions that enable automatic placement and recommendation of video clips for digital products, and 3. Automatically linking and sourcing visual social media content that is relevant to a particular video or online article before it is published. Vidrovr is currently working with these media corporations to develop video centric products and solutions

The Technology

Digital Media meets Computer Vision and Machine Learning

-built and applied towards video understanding-

Index & search multimedia

Search an indexed video corpus and retrieve content inside videos and link it to other media: photos and text.

Extract real entities

Pull out real entities, such as people and organizations out of video streams, articles and images. This is not your old fashioned text search.

Link multimedia content to real news events. See how your and your friends lives are part of the world.

Social integration

Integrating social media, in order to enable direct interactions between multimedia content. Know when the news affects you and your friends.

Connect your APIs

Have a great idea how to use our system-- feel free to connect to our APIs. Use our system's capabilities to solve your problems.

Near realtime

Stream broadcast news in near realtime. We get interactive content to you as quickly as possible, so that you act on the news.

The Team

Joe Ellis


Joe is a PhD candidate at Columbia University working on video searching and indexing. He has published research in the areas of multimodal information processing, computer vision and machine learning. He has worked at Google, IBM Research, and MITRE.

Dan Morozoff


Dan is a PhD candidate at Columbia University working on machine learning applied to neuroscience. He has published research in computational physics, computer multimedia. He has worked at NASA, and HHMI Janelia.

Hongzhi Li

Research Engineer

Hongzhi is a PhD candidate in Computer Science department, Columbia University. His research focuses on Multimedia Cloud Computing and Multimedia Content Analysis. He worked for Microsoft Research and Bell Labs.

Prof. Shih-Fu Chang


Shih-Fu Chang is a professor School of Engineering and at Columbia University. His research focuses on multimedia information retrieval, computer vision, and machine learning with a goal to develop intelligent systems that harness information from massive visual data.


Vidrovr Wins Gold at Publicis 90

July. 2016

Vidrovr is selected as one of the top 12 teams out of 3500+ world applicants to receive funding from Publicis Groupe in their Publicis 90 Startup Competition.
Publicis || TechCrunch || Campaign

Vidrovr presents at NY Meetup

June. 2016

We were invited back to give a short demo of Vidrovr and our tech at the NY Tech Meetup.
NYTM || NYTM-Video

Vidrovr receives followup funding from NYC Media Lab

June. 2016

Following up on 3 months of being part of the Combine. Vidrovr is selected to receive followup funding for the incubation program.

NewsRover Team Won Two Major Challenges in Media Technologies

Feb. 2016

The NewsRover team won two major challenges sponsored by New York City Media Lab. In the Combine program sponsored by NYC Economical Development Commission, the team was selected with major funding to identify products and develop business plans impacting the media industry. As a result, a technology spinoff startup company has been created and attracted additional funding from industry and ventures. In the Verizon challenge, the team won cash prize to further develop prototypes in the new media area.

News Rover @ 2014 NYC Media Lab Annual Summit

Sep 19, 2014

The News Rover system was presented at the 2014 NYC Media Lab Annual Summit and won the second place demo prize. Click here to check the article from NYC Media Lab.

NYC Media Lab

The News Rover team presented the News Rover system @ newsKDD

Aug 24, 2014

The News Rover team presented the News Rover system at newsKDD, a KDD workshop at Bloomberg in New York City.

News Rover @ November 2013 NY Tech Meetup

Nov 4, 2013

The News Rover system was presented at the November 2013 New York Tech Meetup alongside several other projects out from the academic sector at the beautiful NYU Skirball Center For The Performing Arts. It was exciting to present our revolutionary new way of getting context for your news!

Awarded ACM Multimedia 2013 Grand Challenge 1st Place Award

Oct 25, 2013

Work on News Rover presented in part at ACM Multimedia (MM) 2013 won the Multimedia Grand Challenge 1st Place Award in Barcelona, Spain. The work was entitled “Structured Exploration of Who, What, When, and Where in Heterogeneous Multimedia News Sources” and was submitted as a Multimedia Grand Challenge solution to the Technicolor Rich Multimedia Retrieval from Input Videos Grand Challenge. A demo of the work was also presented at MM’13 in a Technical Demo session.
ACM MM 2013