Challenge
Taking part in the development of an AI-based SDK became our first scope in relation to the customer who wanted us to create a tool for neural network self-learning. The more info is accumulated, the higher quality of learning is provided. That is why, the SDK picked up all possible data from the users’ mobile phones, including their time of awakening, movements, and activities. It constantly interacted with backend where these data were analyzed. Neural networks strived to guess which means of transport, bike, car, or bus, the users chose on any particular day or envisage the users’ prevailing emotions. Once the software development kit was ready, the customer addressed us to design a native Android app capable of vividly demonstrating the SDK’s opportunities. That implied a smart music player which detected a person’s mood and played songs to perfectly match and positively impact his/her emotional state.
Approach
According to the requirements, our mobile development team took advantage of the customer’s software development kit to get access to the extensive collection of mood music playlists, motivating text messages, and bright images. In terms of analyzing moments, state of mind, or facial features, we utilized this SDK, too. That contributed to accurate conclusions and lifting people’s spirits on a bad day with the help of Smart Radio.
Description
With the server-side analysis of the aggregated data, the Smart Radio mood music player decides how happy or sad, enthusiastic or exhausted, and sympathetic or furious a person is. Based on the shades of emotions, the mobile app plays the music to suit the users’ mood and current activity and provides them with the appropriate images and motivational quotes.
In case of dynamic exercises and intense workouts, people get heavy melodic playlists to maintain their sporting spirit. If they are tired of physical or mental activities, Smart Radio selects cheerful tracks and shows funny pictures and text motivators to perk up their mood.
Key Features
- The app chooses tracks, background images, and inspirational mottos depending on the user’s mood and activity. Colors and music selection may vary according to the mood swings.
- Smart Radio detects appropriate musical, textual, and visual options from the database located on the server.
- Optionally, the users have a chance to give a cue to the music app on which mood they are in. Also, they are free to delete the tracks they dislike.
Screenshots
Result
Smart Radio is a native mobile app designed by SCAND to reflect and influence the users’ mood through music. It represents the capacities of self-learning neural networks and the AI-driven SDK that can be a key to smart home devices or logistics and medicine tools.
Related Cases
- Java
- Objective-C