Multiple use cases are considered by our client, the final image to be analysed can be produced from a smartphone or an embedded camera into a consumer product. Each use case could generate specific constraints that will require a dedicated dataset. This variety of use case, will generate a broad range of needs in terms of shooting conditions (lens, angle, lighting conditions …). Due to this important need in terms of datasets, the objective is to identify the key partners with the competencies and tools and establish a long-term partnership.
The objective is not only to identify but also to analyse the identified objects or people. We want to be able to measure the physical contours or recognize the facial features of a person and being able to detect the volume, the surface state and other relevant parameters (application dependant) of an object.
The construction of the image dataset and its annotation can be performed separately by two distinct companies or a unique supplier.
Scientific progress has shown the importance of synthetic datasets generation. The datasets can be constructed from real images and enhanced with synthetic datasets. Actors working specifically on the synthetic dataset generation are welcome to apply. For this specific topic, we would be happy to open the discussion with Research Organisation to discuss potential collaboration.
