Leveraging Instance-, Image- and Dataset-Level Information for Weakly Supervised Instance Segmentation
| Autor: | Pei-Song Wen, Yun Liu, Ming-Ming Cheng, Yu-Huan Wu, Yu Qiu, Yujun Shi |
|---|---|
| Rok vydání: | 2022 |
| Předmět: |
Pixel
Computer science business.industry Applied Mathematics Deep learning Pattern recognition 02 engineering and technology Graph Computational Theory and Mathematics Artificial Intelligence Minimum bounding box Computer Science::Computer Vision and Pattern Recognition Prior probability 0202 electrical engineering electronic engineering information engineering Graph (abstract data type) Probability distribution 020201 artificial intelligence & image processing Segmentation Computer Vision and Pattern Recognition Artificial intelligence business Software |
| Zdroj: | IEEE Transactions on Pattern Analysis and Machine Intelligence. 44:1415-1428 |
| ISSN: | 1939-3539 0162-8828 |
| DOI: | 10.1109/tpami.2020.3023152 |
| Popis: | Weakly supervised semantic instance segmentation with only image-level supervision, instead of relying on expensive pixel-wise masks or bounding box annotations, is an important problem to alleviate the data-hungry nature of deep learning. In this article, we tackle this challenging problem by aggregating the image-level information of all training images into a large knowledge graph and exploiting semantic relationships from this graph. Specifically, our effort starts with some generic segment-based object proposals (SOP) without category priors. We propose a multiple instance learning (MIL) framework, which can be trained in an end-to-end manner using training images with image-level labels. For each proposal, this MIL framework can simultaneously compute probability distributions and category-aware semantic features, with which we can formulate a large undirected graph. The category of background is also included in this graph to remove the massive noisy object proposals. An optimal multi-way cut of this graph can thus assign a reliable category label to each proposal. The denoised SOP with assigned category labels can be viewed as pseudo instance segmentation of training images, which are used to train fully supervised models. The proposed approach achieves state-of-the-art performance for both weakly supervised instance segmentation and semantic segmentation. The code is available at https://github.com/yun-liu/LIID. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|