Open Majiawei opened 3 years ago
Thanks for your kind suggestion. For now, we do not have enough resources to do that. PRs are welcomed.
@Majiawei MVXNet is one of 2D-3D fusion methods supported by mmdet3d~
@Majiawei MVXNet is one of 2D-3D fusion methods supported by mmdet3d~
Thank you. I saw it. But I hope to support more 3D object detection methods based on monocular image, binocular image and mult-modal fusion
@Majiawei MVXNet is one of 2D-3D fusion methods supported by mmdet3d~
Thank you. I saw it. But I hope to support more 3D object detection methods based on monocular image, binocular image and mult-modal fusion
Support for monocular 3D will be release in the next few months.
@Majiawei ImVoteNet is also recently added to MMDet3D, which is a multi-modality method for indoor scene 3D object detection.
@Majiawei MVXNet is one of 2D-3D fusion methods supported by mmdet3d~
Thank you. I saw it. But I hope to support more 3D object detection methods based on monocular image, binocular image and mult-modal fusion
Support for monocular 3D will be release in the next few months.
Hope to open d4lcn, am3d, da-3ddet, m3d-rpn and other monocular image based projects as soon as possible, very much looking forward to! The author of d4lcn has open source code, and is also the work of sensetime
Describe the feature As we can see, there are many 3D object detection models based on multi-modal fusion in recent papers. And these models have achieved good results in benchmark dataset, such as mv3d / AVOD / contfuse / MMF / 3d-cvf Moreover, target detection based on multi-modal fusion is one of the trends in the future
Related resources https://github.com/bostondiditeam/MV3D https://github.com/kujason/avod
Additional context Will mmdetection3d support these libraries next? When will it be supported?