Abstract: The present invention relates to a prediction-based technique for encoding light field data by removing redundant information of light field data, reducing a number of bits by employing a prediction of a pixel value in all four dimensions of the light field. Using this technique to represent light field data, allows it to be transferred through a limited-bandwidth medium and/or to significantly reduce the required storage capacity for this purpose.

Abstract: The present invention relates to a prediction-based technique for encoding light field data by removing redundant information of light field data, reducing a number of bits by employing a prediction of a pixel value in all four dimensions of the light field. Using this technique to represent light field data, allows it to be transferred through a limited-bandwidth medium and/or to significantly reduce the required storage capacity for this purpose.

Abstract: A method for creating a non-symmetric codec architecture where a single decoder is able to decode the latent representations produced by different neural encoders. Being a single general decoder, the codec generated does not require multiple symmetric decoders, which saves a large amount of disk space. Therefore, beyond reducing the complexity in execution runtime, the embodiments presented herein significantly reduces the space complexity of learning-based codecs and saves huge amounts of disk space, enabling real applications specially in mobile devices.

Abstract: The present invention relates to a method for compressing light-fields by exploiting their overall 4D redundancy using a hybrid approach that combines the benefits of sparse coding approach with pseudo-video sequence or multi-view coding to explore the inter-SAI redundancy and achieves very competitive results. This redundancy is particularly high when the light-fields are densely sampled. Therefore, this invention is especially efficient for densely sampled light-fields, such as the ones acquired by light-field cameras.

Abstract: A method for compressing light field data using variable block-size four-dimensional transform and bit-plane hexadeca-tree decomposition, the method including: partitioning a four-dimensional pixel data of a light field into four-dimensional blocks of independent fixed size; partitioning the four-dimensional blocks in a set of four-dimensional non-overlapping sub-blocks of different sizes according to a rate-distortion criterion; independently transforming the four-dimensional sub-blocks generated in the previous operation, using a four-dimensional transform; quantizing by bit-planes the sub-blocks transformed by the four-dimensional transform according to Rate-Distortion optimized hexadeca-tree structure; and encoding the four-dimensional quantized data generated through an arithmetic encoder to produce a compressed representation of the light field.