Semantic Instance Segmentation with a Discriminative Loss Function
read the original abstract
Semantic instance segmentation remains a challenging task. In this work we propose to tackle the problem with a discriminative loss function, operating at the pixel level, that encourages a convolutional network to produce a representation of the image that can easily be clustered into instances with a simple post-processing step. The loss function encourages the network to map each pixel to a point in feature space so that pixels belonging to the same instance lie close together while different instances are separated by a wide margin. Our approach of combining an off-the-shelf network with a principled loss function inspired by a metric learning objective is conceptually simple and distinct from recent efforts in instance segmentation. In contrast to previous works, our method does not rely on object proposals or recurrent mechanisms. A key contribution of our work is to demonstrate that such a simple setup without bells and whistles is effective and can perform on par with more complex methods. Moreover, we show that it does not suffer from some of the limitations of the popular detect-and-segment approaches. We achieve competitive performance on the Cityscapes and CVPPP leaf segmentation benchmarks.
This paper has not been read by Pith yet.
Forward citations
Cited by 4 Pith papers
-
Neural Collapse by Design: Learning Class Prototypes on the Hypersphere
Supervised classification is reframed as prototype learning on the hypersphere, with new losses NTCE and NONL achieving neural collapse by design, outperforming cross-entropy on benchmarks and making linear probing un...
-
Neural Collapse by Design: Learning Class Prototypes on the Hypersphere
Supervised classification reaches neural collapse by design via normalized prototype losses on the hypersphere, outperforming CE and SCL on ImageNet-1K and other benchmarks with faster convergence and better transfer.
-
MapRF: Weakly Supervised Online HD Map Construction via NeRF-Guided Self-Training
MapRF reaches about 75% of fully supervised HD map accuracy on Argoverse 2 and nuScenes by generating view-consistent pseudo labels via a NeRF conditioned on map predictions and refining them with Map-to-Ray Matching ...
-
gen2seg: Generative Models Enable Generalizable Instance Segmentation
Finetuning generative models on limited instance segmentation data produces zero-shot generalization to unseen object categories and styles, matching or exceeding supervised baselines like SAM on ambiguous boundaries.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.