Postdoc: Universal INR for Fast Visual Compression

1000scholars; Rennes; On-site; EUR 36,000 - 48,000; Full time

Job summary

1000scholars seeks a postdoctoral researcher to explore universal knowledge extraction from implicit neural representations to cut training time in image compression. The project is a collaboration between Inria and InterDigital under the Nisk.AI umbrella, focusing on efficient neural coding and machine learning for media.

The role emphasizes academic-industrial research with strong theoretical and practical elements.

Responsibilities

• Investigate whether common knowledge can be extracted from image-specific functions to reduce training time.
• Analyze learned INR functions to identify universality across images.
• Develop INR-based compression models with lower encoder cost while retaining benefits of implicit representations.

Job description

1000scholars seeks a postdoctoral researcher to explore universal knowledge extraction from implicit neural representations to cut training time in image compression. The project is a collaboration between Inria and InterDigital under the Nisk.AI umbrella, focusing on efficient neural coding and machine learning for media.

The role emphasizes academic-industrial research with strong theoretical and practical elements.