BroutonLab has worked on a wide range of Data Science projects, from production-ready SaaS to MVPs and prototypes. Check out the list below for a more detailed look into our experience and expertise, and, hopefully, get inspired to add intelligence to your business with us.
Extracting handwritten text from pictures of cheques. Automated the workflow for a bank in Marocco. Reduced the time needed to go through the cheques manually. Achieved the accuracy of 99%.
Detect anomalies (darker or lighter spots) in thermal photos of solar batteries, which are essentially matrixes of panels. OpenCV is used for image preprocessing and extraction of the panel contours. After that, anomaly detection is performed for each of the panels using the Kaze detector.
Predict the outcome of a horse race given the horse's characteristics, its trainer and jockey, as well as race and weather conditions. The pairwise approach is employed for determining the race winner with an XGBoost model. The resulting model achieves a 32% successful prediction rate.
Train a text-to-speech (TTS) model able to generate voice from the given text. Facebook VoiceLoop model is used as the basis with further manual selective layer freezing and retraining for new speakers. The dataset for training is collected from YouTube and further processed with ffmpeg to achieve the best performance while training.
Developed a model to detect explicit content in images. Works in real-time even on mobile devices without a GPU. Has multiple settings to change what is considered NSFW. Outperforms the NSFW model made by Yahoo.
Web-service for medical data visualization and analysis for improving skills of athletes
Talk to us now about your project or a problem you need to solve. Find Data Science solutions to overcome business challenges.
Drop us a line at hello@broutonlab.com or fill the form below.
