At Aksiio, we crack Deep Tech puzzles.
Our team participates in developing an AI-powered 3D body scanning system. We have an exciting opportunity in our team for a strong member with exceptional skills in the field of 3D Computer Vision, Deep Learning on unstructured point cloud data, and human shape and pose estimation.
Responsibilities:
— Develop a deep learning model that predicts optimal human pose and shape parameters directly from raw, noisy depth sensor data;
— Develop optimization techniques for fitting a single set of human shape parameters to a plurarity of point cloud scan data of the same subject in different viewpoints/poses;
— Improve existing synthetic data generation pipelines to generate realistic scan data, mimicking real-world scans with noise and outliers from iPhone 13 and later devices;
— Train and iterate the developed model to achieve high accuracy and robustness on real-world scans;
— Develop a model that is production-ready and can predict avatar parameters in under a minute;
— Deploy the training and inference pipelines on Google Cloud Platform;
— Collaborate with a client engineers to develop cutting-edge technologies in 3D human reconstruction.
Requirements:
— Master’s or PhD in Computer Science or a related field;
— At least 3 years of relevant industry experience in 3D computer vision, deep learning, and training and deploying new machine learning models for human shape and pose estimation;
— Strong proficiency in Python and relevant deep learning frameworks (e.g., TensorFlow, PyTorch, Keras);
— Prior experience with deep learning models for point cloud data;
— Prior experience with statistical human body models such as SMPL/STAR;
— Prior experience with deep learning models for point cloud data, such
— Prior experience with iPhone TrueDepth camera, Kinect sensors, or other structured light sensors;
— Prior experience with training, iterating, and deploying models on Google Cloud Platform.
If you are passionate about developing cutting-edge technologies in 3D human reconstruction, we encourage you to apply.