talk-data.com

The project aims to provide a user-friendly web application that empowers users to assess the risk of cervical cancer and offers the ability to analyze images for potential tumor presence. By leveraging machine learning techniques and incorporating XAI and text generation, the application can contribute to early detection, promoting timely medical intervention.

We are developing a LLM that uses medical question-answer pairs from PubMedQA: Jin et al., 2019, MedQA: Jin et al., 2020, and MMLU: Hendrycks et al., 2020. This model is mainly aimed as an instructional tool for non-EU doctors who want to become licensed doctors in Germany.

The first is recording patient-doctor consultations: transcription using audio-to-text models, and then summarising and reformatting into professional clinical documentation using LLMs. The second is using NLP and vector databases to do efficient search of clinical guideline documents, and ability to interact with the guidelines using NLP.

Love cycling, hate cobblestones? This project uses recent street view images of Berlin to help people plot the best bicycle route from A to B.

Leveraging Computer Vision for People Tracking and Object Detection. The application tracks and analyze people’s movements in various environments such as retail stores, shopping malls, warehouses, public spaces, airports, etc The Key Features: count people, create heatmaps, detect queues, track movement patterns, space occupancy analysis, zone-specific analytics, and customer movement patterns, providing valuable insights.

NaturalMaps is an attempt to explore maps with natural language, such as “Find all the quietest coffee shops in Berlin that open before 8 AM and are in close proximity to a library.” At the moment, this is readily accessible open data, but going beyond simple queries requires expert know-how. We are exploring ways to make this information and analysis more accessible to the user