Learning Dense Convolutional Embeddings for Semantic Segmentation

Abstract : This paper proposes a new deep convolutional neural network (DCNN) architecture that learns pixel embeddings, such that pairwise distances between the embeddings can be used to infer whether or not the pixels lie on the same region. That is, for any two pixels on the same object, the embeddings are trained to be similar; for any pair that straddles an object boundary, the embeddings are trained to be dissimilar. Experimental results show that when this embedding network is used in conjunction with a DCNN trained on semantic segmentation, there is a systematic improvement in per-pixel classification accuracy. Our contributions are integrated in the popular Caffe deep learning framework, and consist in straightforward modifications to convolution routines. As such, they can be exploited for any task involving convolution layers.
Document type :
Preprints, Working Papers, ...
Liste complète des métadonnées

https://hal.inria.fr/hal-01263620
Contributor : Iasonas Kokkinos <>
Submitted on : Wednesday, January 27, 2016 - 10:12:11 PM
Last modification on : Tuesday, February 5, 2019 - 1:52:14 PM

Links full text

Identifiers

  • HAL Id : hal-01263620, version 1
  • ARXIV : 1511.04377

Citation

Harley Adam, Iasonas Kokkinos, Kostas Derpanis. Learning Dense Convolutional Embeddings for Semantic Segmentation . 2015. ⟨hal-01263620⟩

Share

Metrics

Record views

241