All of us know what happens after the camera's are rolled and the actions that followed. How many of us are aware of things that happen before a stage is set, lines memorised and actors are chosen ?

A typical Audition take about 5-10 minutes per person and a report claims that for a single person it may take between 150-200 auditions to land a role so thats 150 audition per person** and there are often 100s if not 1000s auditioning for a chance to act in their fav moive or a film.

The hours taken to audit them is tedious, this weekend we wanted to do our bit to save as many hours as possible with the help of Wix Velo and Open CV

What it does

Our Talent Scour platform Hactor is designed to make finding budding actors easier, acting in drams, movies or even on a silent movie is emotionally demanding. An actor must potray a range of emotions at times without a word

Our job applicants will be give a line to act on with sometimes a emotion The applicant then has to enact this line in terms of the given emotion and upload a mp4 veideo on the application. This website is created on Wix. Velo is responsible in communicating with our backend api for data transfer , storage and analytics. As soon as the user submits his application , the async ML API is triggered with the Veideo link for the processing to start. The processing time is linear to the size and frame count of the veideo. Once the procesing is complete the ML api then return a json response with the detected emotions as a percentage and the user id back to our JS api which then stores it on DB for later use This score is shown to the Production house / director that posts this job This can help the recuiters understand the applicants performance to a particular intent and can decide to move on based on the result .

How we built it

We decided to build our UI on WIX and with Velo we had the oppurtunity to make it dynmaic and interactable with our API The backed API is built on JS The ML api is built on Flask and uses Open CV , containerized using Docker Deployed on Azure

Challenges we ran into

Deploying the OpenCV app to azure was a challange on its own , it took us almost half a day just to get it ready , we were met with a lack of documentation and guide for this usecase and needless a new article is born out of this hackathon hoping to share it with you all soon.

Accomplishments that we're proud of

The ML API is something we didnt expect to perform so well, since its a short hackathon and we havnt worked with wix and open cv before it was a challanging undertaking. We were able to build this and bring it to you as a demo and thats something we are proud of

What we learned

Wix is something we learned as a team , needless to say we will be exploring it more over the next weekend as well

What's next for Hactor

We want to build this app full time and interact with production houses to know more about additional requirements


Link to Github

Built With

Share this project: