File tree Expand file tree Collapse file tree 1 file changed +1
-1
lines changed
Expand file tree Collapse file tree 1 file changed +1
-1
lines changed Original file line number Diff line number Diff line change @@ -196,7 +196,7 @@ <h2 class="title is-3">Method</h2>
196196 < p >
197197 < b > Shared Mapping is the key to bridge generation and understanding ability in one tokenizer for AR model.</ b >
198198 We incorporate dual encoders and codebooks with a shared mapping, enabling the joint optimization of high-level
199- semantics and low-level pixel details. For a given input image, distances are calculated from the pixel-level land semantic-level codebooks,
199+ semantics and low-level pixel details. For a given input image, distances are calculated from the pixel-level and semantic-level codebooks,
200200 respectively, with the final codebook index and features determined by minimizing the weighted sum distance.
201201 The resulting quantized features are independently decoded for both semantic alignment and image reconstruction training,
202202 and then concatenated to provide a unified representation for downstream tasks in understanding and generation.
You can’t perform that action at this time.
0 commit comments