.png){kind=icon}
.png){kind=icon}
Return
QuantizationGS: Differentiable quantization model for 3D Gaussian Splatting compression
.png){kind=icon}
PRE
AI
DOI：10.1016/j.cag.2026.104596.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• Compress the model by injecting quantization noise into the parameters during training. • Prune redundant anchor points using a global importance scoring mechanism to improve rendering speed. • Maintain reconstruction quality through specialized loss constraining model training.
Keywords:
Quantization
Gaussian Splatting
Model Compression
Anchor Point Pruning
Loss Constraining
Journal
C
IF:
0
Papers:
119
Citations:
0