Image and Video Tokenization with Binary Spherical Quantization.

Yue Zhao Yuanjun Xiong Philipp Krähenbühl

Published in: CoRR (2024)

Keyphrases

image data
single image
multiscale
input image
image representation
image content
image classification
image segmentation
image frames
video streams
low level
image retrieval
image features
multimedia
edge detection
segmentation method
key frames
image collections
quantization error
jpeg images
quantization noise
temporal continuity
uniform quantization