GAP-MLLM: Geometry-Aligned Pre-training for Activating 3D Spatial Perception in Multimodal Large Language Models
Open paper
Benchmark
Model
Rank
Results
3d-dense-captioning-on-scan2cap
GAP-MLLM-3B
#1
CIDEr: 84.7
ROUGE: 63.1
Rank counts only results with a code link.