Papers › AdaptIS: Adaptive Instance Selection Network

AdaptIS: Adaptive Instance Selection Network

17 Sep 2019ICCV 2019 10arXiv:1909.07829archive 2025-07-28

Konstantin Sofiiuk, Olga Barinova, Anton Konushin

We present Adaptive Instance Selection network architecture for class-agnostic instance segmentation. Given an input image and a point (x, y), it generates a mask for the object located at (x, y). The network adapts to the input point with a help of AdaIN layers, thus producing different masks for different objects on the same image. AdaptIS generates pixel-accurate object masks, therefore it accurately segments objects of complex shape or severely occluded ones. AdaptIS can be easily combined with standard semantic segmentation pipeline to perform panoptic segmentation. To illustrate the idea, we perform experiments on a challenging toy problem with difficult occlusions. Then we extensively evaluate the method on panoptic segmentation benchmarks. We obtain state-of-the-art results on Cityscapes and Mapillary even without pretraining on COCO, and show competitive results on a challenging COCO dataset. The source code of the method and the trained models are available at https://github.com/saic-vul/adaptis.

PaperPDFConference PDF

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

No code repository is listed for this paper in the archive or in Syntology's graph.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Instance SegmentationPanoptic SegmentationSegmentationSemantic Segmentation

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Panoptic Segmentation COCO test-dev AdaptIS (ResNeXt-101) PQ 42.8 #30 of 38 Archive leaderboard report
Panoptic Segmentation COCO test-dev AdaptIS (ResNeXt-101) PQst 31.8 #30 of 38 Archive leaderboard report
Panoptic Segmentation COCO test-dev AdaptIS (ResNeXt-101) PQth 50.1 #30 of 38 Archive leaderboard report
Panoptic Segmentation Cityscapes val AdaptIS (ResNeXt-101) AP 36.3 #21 of 37 Archive leaderboard report
Panoptic Segmentation Cityscapes val AdaptIS (ResNeXt-101) PQ 62.0 #21 of 37 Archive leaderboard report
Panoptic Segmentation Cityscapes val AdaptIS (ResNeXt-101) PQst 64.4 #21 of 37 Archive leaderboard report
Panoptic Segmentation Cityscapes val AdaptIS (ResNeXt-101) PQth 58.7 #21 of 37 Archive leaderboard report
Panoptic Segmentation Cityscapes val AdaptIS (ResNeXt-101) mIoU 79.2 #21 of 37 Archive leaderboard report
Panoptic Segmentation Cityscapes val AdaptIS (ResNet-101) AP 33.9 #25 of 37 Archive leaderboard report
Panoptic Segmentation Cityscapes val AdaptIS (ResNet-101) PQ 60.6 #25 of 37 Archive leaderboard report
Panoptic Segmentation Cityscapes val AdaptIS (ResNet-101) PQst 62.9 #25 of 37 Archive leaderboard report
Panoptic Segmentation Cityscapes val AdaptIS (ResNet-101) PQth 57.5 #25 of 37 Archive leaderboard report
Panoptic Segmentation Cityscapes val AdaptIS (ResNet-101) mIoU 77.2 #25 of 37 Archive leaderboard report
Panoptic Segmentation Cityscapes val AdaptIS (ResNet-50) AP 32.3 #31 of 37 Archive leaderboard report
Panoptic Segmentation Cityscapes val AdaptIS (ResNet-50) PQ 59.0 #31 of 37 Archive leaderboard report
Panoptic Segmentation Cityscapes val AdaptIS (ResNet-50) PQst 61.3 #31 of 37 Archive leaderboard report
Panoptic Segmentation Cityscapes val AdaptIS (ResNet-50) PQth 55.8 #31 of 37 Archive leaderboard report
Panoptic Segmentation Cityscapes val AdaptIS (ResNet-50) mIoU 75.3 #31 of 37 Archive leaderboard report
Panoptic Segmentation Mapillary val AdaptIS (ResNeXt-101) PQ 40.3 #9 of 13 Archive leaderboard report
Panoptic Segmentation Mapillary val AdaptIS (ResNeXt-101) mIoU 56.8 #9 of 13 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

1x1 ConvolutionAverage PoolingBatch NormalizationBottleneck Residual BlockConvolutionGlobal Average PoolingGrouped ConvolutionKaiming InitializationMax PoolingReLUResNeXtResNeXt BlockResidual BlockResidual Connection

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections