Post-doctoral researcher; Hybrid Implicit Neural Representations for Next-Generation Learned Video Compression
Core
Develop a new generation of learned video codecs based on spatio-temporal hybrid implicit neural representations to reduce the environmental footprint of generative AI.
Role type
Post-doctoral researcher (research & standards)
Builds
Learned video compression frameworks and international standards
Domain
AI research, video compression, wireless technology
Deliverable
research
Required skills
Implicit Neural Representations (INRs), learned video compression, hierarchical latent modeling, entropy modeling, neural architecture design
Preferred skills
spatio-temporal modeling, continuous temporal prediction, lightweight neural networks
Technologies
PyTorch, TensorFlow, C++, CUDA
Responsibilities
Design novel spatio-temporal hybrid INR architectures, develop lightweight decoder architectures, conduct rate-distortion-complexity evaluation against state-of-the-art codecs
Seniority
Post-doctoral researcher
