Success Stories
Since our inception, HSA has collaborated with a diverse range of customers, each with unique challenges and objectives. Our team of experts has consistently provided customized solutions that not only meet but exceed expectations. Through our innovative approach and unwavering commitment to quality, we have established ourselves as a trusted partner in the industry.
Here are some of our success stories:
Drone Footage Analysis
Our customer utilized HSA KIT to analyze multispectral drone footage of fields, focusing on weed count and crop health. They deployed HSA KIT on their own servers, leveraging an optimized Kubernetes (K8s) cluster and pulling updates from the HSA AWS ECR repository. This setup enabled a distributed team of international annotators to generate ground truth data for further model training.
- Key HSA KIT Benefits:
- Intuitive Interface: Easy navigation for annotators.
- High Multichannel Performance: Smooth and lag-free operation even with multispectral channels aside from RGB.
- Optimized Data Display: Detailed multichannel visualization.
- Advanced Histogram Tools: Precise data analysis.
- Distributed Annotation:
- The deployment allowed for a distributed team of international annotators to collaborate seamlessly, leveraging the intuitive and performant features of the HSA KIT.
- Ground Truth Data Generation:
- Annotators used the HSA KIT to generate accurate ground truth data from drone footage, essential for training machine learning models aimed at improving crop and weed detection algorithms.
- Efficiency and Scalability:
- The use of a Kubernetes cluster ensured that the annotation environment was both scalable and efficient, handling the demands of high-volume data processing and annotation tasks with ease.
Zebrafish Analysis
The deployment of HSA KIT in AWS demonstrated remarkable success in the analysis of zebrafish heart videos. The process involved:
- Video Import and Processing:
- Videos were imported and converted into sequences for analysis.
- Manual selection of Regions of Interest (ROI) was followed by automated segmentation using deep learning.
- Heart parameters such as rhythm, frequency, and fractional shortening were accurately measured.
- Cloud Benefits:
- The cloud setup enabled rapid and scalable processing of large video datasets.
- Researchers could access and analyze data remotely, significantly enhancing collaborative efforts.
