ProGIC: Progressive and Lightweight Generative Image Compression with Residual Vector Quantization
Researchers have developed ProGIC, a new generative image compression method that uses residual vector quantization for progressive, lightweight compression. This approach allows for a coarse-to-fine reconstruction and a progressive bitstream, enabling previews from partial data. ProGIC achieves comparable compression performance to existing methods while offering significant improvements in speed and efficiency, making it suitable for practical deployment on various devices. AI
IMPACT Introduces a more efficient and flexible image compression technique suitable for real-world applications.