In our projects we apply swarmcreativity to increase happiness of living beings. This means to improve connectivity between humans, animals, and plants. In our projects we will form COINs (Collaborative Innovation Networks) to create tools to measure and improve communication between humans, humans and animals, and humans and plants.
See www.biolingo.org for an overview.
Connecing with the right people, sharing your values and interests, will increase your groupflow and happiness. Projects in this field are using AI to measure your happiness from your emotions measured from text, smartphones, face expressions, voice, and body posture.
Connecting people with animals to understand each other and communicate better will increase both the happiness and wellbeing of your pets, such as dogs and cats, and horses and cows and yourself. Projects in this domain will be using videos, voice, and other sensor data tracking animals to compare it to ground truth such as audio and image files labeled with the emotions and health status of animals. We will also be using unsupervised learning to further identify mood states of animals that we do not know yet.
Connecting with nature is good for you. Projects in this area will help you communicate with your house plants. In past work we have developed specialized hardware that measures the "brainwaves" of plants, similar to EEGs of human brains. These "plant action potentials" are then fed as input to machine learning models to predict human mood. We will for instance use plants such as basil as sensors to measure human emotions and stress in a non-intrusive way.
Correlating animal movement patterns with natural disasters
Using animal movement data from sources such as Movebank (https://datarepository.movebank.org/home) and disaster event records from databases like USGS (https://www.usgs.gov) or EM-DAT (https://www.emdat.be), conduct a time-series analysis to investigate whether animals show significant behavioral changes before or during natural disasters, extending prior work on animal-based early warning systems.
Personality differences among professions
Using pretrained models recognizing personality and morals from text, collect a dataset of different professions from LinkedIn and other social media sources. Then aggregate the text produced by members of different professions, and identify personality characteristics typical of professions. The same approach could even be applied to automated face analysis, e.g. using models that predict facial-width-to-height ratio.
Recognizing pain in cows
Using a video dataset collected by a veterinary professor in Spain, build an automated app to recognize if a cow is having pain.
Do entangled Jazz musicians play better
Using a dataset from the Jazzaar festival 2023 consisting of videos, measure synchronization among musicians and strenght of applause. Are more synchronized musicians playing better - if so, how is synchronization happening. An earlier COIN seminar team already explored synchronicity of musicians in rehearsals based on face and voice analysis.
Do plants like Jazz?
Using the same dataset investigate the correlation between the plant electrical signals recorded with a plant spikerbox from basils placed in the vicinity of musicians and audience and the face and body synchronicity of the Jazz musicians as well as with the audience.
Do entangled teams collaborate better?
Using a dataset collected in a one week course at MIT among 90 students collaborating in 22 teams, we will do video analysis using body posture video analysis software to measure entanglement among team members. Each team (4-5 students) sat around a table equipped with video camera and microphone.
Are plants capable of measuring collaboration quality?
In the same teamwork setting as just described above, a basil plant on each table equipped with a plant spikerbox recorded voltage differences of the plant. These will be compared with emotions of participants computed from the videos.