Skip to article frontmatterSkip to article content

LLaMARJO webapp

Livestock labeling Learning for Manual Analysis & Recognition as a Journal of Observations webappยถ

An example screenshot of the LLaMARJO webapp

An example screenshot of the LLaMARJO webapp

๐ŸŽฏ Project objectiveยถ

Ease the learning and evaluation process of behavior annotation in videos both for the teacher and the student.

๐Ÿ“‹ Project descriptionยถ

LLMARJO is a software that enables user-centric learning in the processes of behavior annotation.

The process for training in LLAMARJO is similar to machine learning, meaning that three datasets are defined per user for training, validating and testing their skills. The datasets were previously annotated by domain experts and annotations are supplied only for training data to the student.

After having created a learning course for the student, the teacher is there to provide feedback but doesnโ€™t need to accompany the student in the whole process. The learning process is handled by the student enabling them to learn, practice and get certified by the system at their own pace. Afterwards, the teacher only need to confirm the evaluations before accrediting the students for production tasks.

๐ŸŽจ Design & implementation decisionsยถ

๐Ÿงพ Key takeawayยถ

๐Ÿ‘จโ€๐Ÿ’ป Contribution:ยถ

  • Global design

  • Data preprocessing

  • Database building

  • Data storage in Google Drive

๐Ÿ›  Tools:ยถ

  • Google AppSheet

  • Python

  • Pandas

  • FFmpeg

  • Google Drive API

  • Youtube API

  • Speech T5 by Microsoft from Hugging Face Hub

Github repository