Multispectral Object Detection On Flir 1

Metrics

mAP

mAP50

Results

Performance results of various models on this benchmark

			Paper Title
MMPedestron	-	86.4%	When Pedestrian Detection Meets Multi-Modal Learning: Generalist Model and Benchmark Dataset
RGB-X Scene Adaptive CBAM	47.1%	86.16%	RGB-X Object Detection via Scene-Specific Fusion Modules
CAFF-DINO	50.5%	85.5%	-
RSDet	43.8%	83.9%	Removal then Selection: A Coarse-to-Fine Fusion Perspective for RGB-Infrared Object Detection
CMX	-	82.2%	CMX: Cross-Modal Fusion for RGB-X Semantic Segmentation with Transformers
UniRGB-IR	44.1%	81.4%	UniRGB-IR: A Unified Framework for Visible-Infrared Semantic Tasks via Adapter Tuning
MiPa	44.8%	81.3%	MiPa: Mixed Patch Infrared-Visible Modality Agnostic Object Detection
CSSA	41.3%	79.2%	Multimodal Object Detection by Channel Switching and Spatial Attention
CFT	-	77.7%	Cross-Modality Fusion Transformer for Multispectral Object Detection
ProbEn	37.9%	75.5%	Multimodal Object Detection by Channel Switching and Spatial Attention
GAFF	37.4%	74.6%	Multimodal Object Detection by Channel Switching and Spatial Attention
YOLOv5 (T)	-	73.9%	Cross-Modality Fusion Transformer for Multispectral Object Detection
GAFF (ResNet18)	-	72.9%	Guided Attentive Feature Fusion for Multispectral Pedestrian Detection
GAFF (VGG16)	-	72.7%	Guided Attentive Feature Fusion for Multispectral Pedestrian Detection
CFR_3 (VGG16)	-	72.4%	Multispectral Fusion for Object Detection with Cyclic Fuse-and-Refine Blocks
Halfway Fusion (VGG16)	-	71.2%	Multispectral Fusion for Object Detection with Cyclic Fuse-and-Refine Blocks
YOLOv5 (RGB)	-	67.8%	Cross-Modality Fusion Transformer for Multispectral Object Detection
Halfway Fusion	35.8%	-	Multimodal Object Detection by Channel Switching and Spatial Attention

0 of 18 row(s) selected.

Multispectral Object Detection On Flir 1 | SOTA | HyperAI