The challenge is to develop a real-time object detection model for autonomous vehicles using computer vision techniques and Intel® AI Analytics Toolkits its libraries, and SYCL/DPC++ Libraries . Participants are expected to create a deep learning model that can accurately detect objects such as pedestrians, vehicles, traffic signs, and traffic signals. The model should perform with high accuracy and low latency, ensuring safe navigation for autonomous vehicles.

End Goal

• Develop a deep learning model using Intel® AI Analytics Toolkits its libraries, and SYCL/DPC++ Libraries to detect objects on the road, with a focus on accuracy and speed.

• Test the model on a dataset that includes real-world scenarios, including various weather conditions, lighting conditions, and road environments.

• The solution should be able to prevent load shedding by providing timely alerts to energy providers to take necessary action to prevent power grid overload.

• Optimize the model for low latency to enable real-time object detection in autonomous vehicles.

• Provide a comprehensive report detailing the approach, methodology, results, and challenges faced during the development and testing of the model.

  Note: It is mandatory to use the Intel® AI Analytics Toolkits, its Libraries and SYCL/DPC++ Libraries for the solution. Leverage the libraries for maximum efficiency...

Processing medical images can be time-consuming, especially when dealing with large datasets. Therefore, the challenge is to optimise medical image processing using SYCL to improve the accuracy and speed of diagnosis.

End Goal

The end goal is to build a medical image processing application that demonstrates the performance benefits of using SYCL for parallel computing. The application should process medical images faster and with higher accuracy than existing medical imaging applications.

• Use SYCL's parallelism capabilities to speed up medical image processing.

• Improve the accuracy of diagnosis through efficient feature extraction from medical images.

• Benchmark the application against existing medical imaging applications in terms of performance, accuracy, and scalability.

• Use a publicly available medical imaging dataset, such as the NIH Chest X-ray dataset or the BraTS dataset for brain tumour segmentation.

The education sector is facing numerous challenges in the wake of the COVID-19 pandemic. One of the major challenges is the need for innovative solutions to address the disparities in student learning outcomes and to ensure equitable access to quality education. To tackle this challenge, we are calling for an open innovation challenge that leverages the Intel® AI Analytics Toolkits, its libraries and the SYCL/DCP++ Libraries to develop innovative solutions that can improve student learning outcomes and promote equitable access to quality education.

End Goal

• The ultimate goal of this open innovation challenge is to create innovative solutions that leverage the Intel® AI Analytics Toolkits its libraries and the SYCL/DCP++ Libraries to improve student learning outcomes and promote equitable access to quality education.

Note: To achieve optimal performance and better models, it is required to use the Intel® AI Analytics Toolkits its libraries and the SYCL/DCP++ Libraries in the development of the interactive educational platform. Therefore, participants are advised to leverage the libraries and optimizations provided by oneAPI to enhance the efficiency of the application.