Netra Sponsors “Hack the Pack” Hackathon at Northeastern University
Recently Netra’s co-op for the Research and Development team, Oliver, headed an effort to get Netra involved in the hackathon scene. As he is a student at Northeastern University, he connected us to a Northeastern student hackathon, Hack the Pack, which Netra ended up sponsoring. This was Netra’s first time sponsoring a hackathon, and it turned out to be a great success. Netra’s main goal was to see how easy we could make it for someone to get their hands on our API, and test it using their own data. To this end, Oliver created and ran a 1 hour workshop on how to use Netra’s API, and how to analyze the data obtained through it.
Content
For this workshop, we decided to show how our API can be used to analyze Northeastern students, or huskies. The code and slides for this workshop can be found here.
The first step was to collect publicly available pictures from social profiles of huskies. We collected from a few hundred profiles, which included both current students and Northeastern Alumni. Next, we set up the environment in which the attendees would send the images through the API. To do this, we wrote a script to send the images through the API, and a small flask server to receive the callbacks. Once the flask server was running locally, we used ngrok to tunnel requests to it, and then sent the images through the API. For the workshop the students received a small subset of the images, and we pre-processed the rest of the images for them for analysis.
Once the images had been processed, the next step was to process the information received from the API. For the workshop, we showed how this could be done with `python`, but this could be done in many environments. The first step was to anonymize the URLs. We had already anonymized the usernames, but thought it would be helpful to show how anonymization works via the URLs. After processing the JSON and converting it to a csv format programatically, we begun the analysis of the results.
Through the Brands API, we found that some of the most popular brands among huskies included Nike, Adidas, and North Face. This indicates that Northeastern students are an outdoorsy group. This result was confirmed by our Context API, as some of the common tags included ‘Body of Water’, ‘Terrain’, and ‘Flowers, Plants, Trees’. Overall, the information we got from our APIs seemed to give us some great insight into what types of people go to Northeastern.
The Workshop
About 15–20 Northeastern students showed up to the workshop, which was a good turnout. For the purpose of timeliness, we decided to let the students know they didn’t have to follow along with the code on their own, but strongly encouraged it. The event turned out to be a great way for Netra to get its feet wet in the hackathon scene. Like any first attempt though, it was definitely a learning experience. One of the main issues we found was that completing the workshop took quite a bit of setup, as students needed to install `jupyter`, `flask`, `ngrok`, and any missing python libraries. This can be too high of a barrier to entry to beginning coders, and we’d like to make setup easier in the future.
Final Thoughts
Overall, this was a very positive first experience for Netra in the hackathon scene. We were very happy to help new students get involved in using our API, and may be looking to do a larger event in the future. A side benefit of this is that it has given us a launching point for improving our ability to get people set up with our API quickly. We’d like to give a big shout out to the organizers of Hack the Pack for running such a great event. Make sure to visit their event site here.
Netra develops image and video recognition APIs to help enterprise structure and make sense of their visual media. Netra’s API ingests photo or video URLs and, within milliseconds, automatically tags it for visual content such as brand logos, objects, scenes, and people with demographic classification. If you’re interested in learning more, visit our website or say hello at info@netra.io !
