Enhanced Multi-Instance Partial Label Learning via Average Gradient Outer Product

Nan Cao, Xu Zhao, Teng Zhang
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:11830-11842, 2026.

Abstract

Multi-instance partial-label learning (MIPL) is a recently proposed dual weakly supervised learning framework where each training bag is annotated with a candidate label set containing one true label and several false positives. The key challange of MIPL problem is that the relations between the key instances and the ground-truth labels are much more covered due to the false positive labels. Existing methods usually rely on the model’s own predictions and roughly aggregate instance features according to the learned attention weights, which is easily misled by false positives and lets key instances be overlooked. Here, we propose Average Gradient Outer Product based Multi-instance Partial-Label Learning (AGOPMIPL) method, where the average gradient outer product (AGOP) is directly calculated from the bag-level features and the model prediction, then it is integrated in an attention module to amplify discriminative feature directions and thereby helps key-instance identification. Moreover, the feature prototypes and a progressive disambiguation strategy are introduced to further suppress noisy candidates. The experimental studies on four MIPL benchmarks and the real-world CRC-MIPL dataset are performed and AGOPMIPL consistently outperforms five state-of-the-art baselines, with up to $25.9%$ relative gain on CRC-MIPL-KMeansSeg.

Cite this Paper

BibTeX
@InProceedings{pmlr-v306-cao26af, title = {Enhanced Multi-Instance Partial Label Learning via Average Gradient Outer Product}, author = {Cao, Nan and Zhao, Xu and Zhang, Teng}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {11830--11842}, year = {2026}, editor = {Zhang, Tong and Dudik, Miroslav and Jaggi, Martin and Agarwal, Alekh and Li, Sharon and Schuurmans, Dale and Zhu, Jerry and Berkenkamp, Felix and Dong, Hanze and Bietti, Alberto}, volume = {306}, series = {Proceedings of Machine Learning Research}, month = {06--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v306/main/assets/cao26af/cao26af.pdf}, url = {https://proceedings.mlr.press/v306/cao26af.html}, abstract = {Multi-instance partial-label learning (MIPL) is a recently proposed dual weakly supervised learning framework where each training bag is annotated with a candidate label set containing one true label and several false positives. The key challange of MIPL problem is that the relations between the key instances and the ground-truth labels are much more covered due to the false positive labels. Existing methods usually rely on the model’s own predictions and roughly aggregate instance features according to the learned attention weights, which is easily misled by false positives and lets key instances be overlooked. Here, we propose Average Gradient Outer Product based Multi-instance Partial-Label Learning (AGOPMIPL) method, where the average gradient outer product (AGOP) is directly calculated from the bag-level features and the model prediction, then it is integrated in an attention module to amplify discriminative feature directions and thereby helps key-instance identification. Moreover, the feature prototypes and a progressive disambiguation strategy are introduced to further suppress noisy candidates. The experimental studies on four MIPL benchmarks and the real-world CRC-MIPL dataset are performed and AGOPMIPL consistently outperforms five state-of-the-art baselines, with up to $25.9%$ relative gain on CRC-MIPL-KMeansSeg.} }
Endnote
%0 Conference Paper %T Enhanced Multi-Instance Partial Label Learning via Average Gradient Outer Product %A Nan Cao %A Xu Zhao %A Teng Zhang %B Proceedings of the 43rd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2026 %E Tong Zhang %E Miroslav Dudik %E Martin Jaggi %E Alekh Agarwal %E Sharon Li %E Dale Schuurmans %E Jerry Zhu %E Felix Berkenkamp %E Hanze Dong %E Alberto Bietti %F pmlr-v306-cao26af %I PMLR %P 11830--11842 %U https://proceedings.mlr.press/v306/cao26af.html %V 306 %X Multi-instance partial-label learning (MIPL) is a recently proposed dual weakly supervised learning framework where each training bag is annotated with a candidate label set containing one true label and several false positives. The key challange of MIPL problem is that the relations between the key instances and the ground-truth labels are much more covered due to the false positive labels. Existing methods usually rely on the model’s own predictions and roughly aggregate instance features according to the learned attention weights, which is easily misled by false positives and lets key instances be overlooked. Here, we propose Average Gradient Outer Product based Multi-instance Partial-Label Learning (AGOPMIPL) method, where the average gradient outer product (AGOP) is directly calculated from the bag-level features and the model prediction, then it is integrated in an attention module to amplify discriminative feature directions and thereby helps key-instance identification. Moreover, the feature prototypes and a progressive disambiguation strategy are introduced to further suppress noisy candidates. The experimental studies on four MIPL benchmarks and the real-world CRC-MIPL dataset are performed and AGOPMIPL consistently outperforms five state-of-the-art baselines, with up to $25.9%$ relative gain on CRC-MIPL-KMeansSeg.
APA
Cao, N., Zhao, X. & Zhang, T.. (2026). Enhanced Multi-Instance Partial Label Learning via Average Gradient Outer Product. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:11830-11842 Available from https://proceedings.mlr.press/v306/cao26af.html.

Related Material