Comparer des méthodes

	Segmentation d'instances multimodale	Détection d'objets multimodale
Domaine	Apprentissage profond	Apprentissage profond
Famille	Machine learning	Machine learning
Année d'origine	2017–present	2015–2019
Auteur d'origine	He, K., Gkioxari, G., Dollar, P., Girshick, R. (Mask R-CNN foundation); extended by community to multimodal settings	Multiple contributors (e.g., Chen & Deng, Liang et al.)
Type	Supervised deep learning — instance segmentation	Fusion-based deep detection
Source fondatrice	He, K., Gkioxari, G., Dollar, P., & Girshick, R. (2017). Mask R-CNN. Proceedings of the IEEE International Conference on Computer Vision (ICCV), 2961–2969.	Liu, Y., Zhang, F., Li, Y., & Lv, H. (2022). Multimodal Object Detection via Bayesian Fusion. IEEE Transactions on Image Processing, 31, 5953–5965.
Alias	multimodal Mask R-CNN, RGB-D instance segmentation, multi-sensor instance segmentation, cross-modal instance segmentation	multi-sensor object detection, cross-modal detection, RGB-D object detection, fusion-based object detection
Apparentées	5	6
Résumé	Multimodal instance segmentation extends classical instance segmentation — which assigns a per-pixel mask and a class label to every individual object in an image — by incorporating complementary sensor streams such as depth maps, LiDAR point clouds, or infrared frames. Fusing these modalities helps the model handle ambiguous appearances, low light, and occlusion that trip up RGB-only systems.	Multimodal object detection extends single-modality object detectors by jointly processing signals from multiple sensor types — such as RGB cameras, depth sensors, LiDAR, radar, or text descriptions — to localize and classify objects with higher accuracy and robustness than any single modality alone. Fusion of complementary information is the core design principle.