Research map / Detection and segmentation / Open-set and few-shot detection
Unsupervised Learning of Foreground Object Segmentation
Ioana Croitoru, Simion-Vlad Bogolin, Marius Leordeanu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| run a second student-teacher iteration | LowRes-Net · YouTube Objects v1 | CorLoc: 62.1 -> 63.5 | Improved |
| run a second student-teacher iteration | FConv-Net · YouTube Objects v1 | CorLoc: 57.6 -> 61.3 | Improved |
| run a second student-teacher iteration | BasicU-Net · YouTube Objects v1 | CorLoc: 59.1 -> 65.2 | Improved |
| run a second student-teacher iteration | MultiSelect-Net · YouTube Objects v1 | CorLoc: 62.4 -> 67.3 | Improved |
| run a second student-teacher iteration | single models average · Object Discovery in Internet Images | CorLoc: 84.3 -> 86.8 | Improved |
| run a second student-teacher iteration | FConv-Net · Object Discovery in Internet Images | CorLoc: 79.8 -> 85.6 | Improved |
| run a second student-teacher iteration | FConv-Net · PASCAL-S | F-beta: 51.5 -> 61.7 | Improved |
| ensemble several student nets (Multi-Net) vs single average | student nets · YouTube Objects v1, iteration 2 | CorLoc: 63.0 -> 67.2 | Improved |
| add unlabeled YTO data to VID for training (ablation) | LowRes-Net · YouTube Objects v1 | CorLoc: 56.1 -> 62.2 | Improved |
| add unlabeled YTO data to VID for training (ablation) | LowRes-Net · Object Discovery in Internet Images | mean J: 61.25 -> 62.33 | Improved |
| add unlabeled YTO data to VID for training (ablation) | LowRes-Net · Object Discovery in Internet Images | mean P: 87.73 -> 88.36 | Improved |
| crop image by LowRes-Net foreground box before VGG19 features, combined with whole image (ablation) | VGG19 classifier · YouTube Objects v1 classification | recognition rate: 69.1 -> 72.7 | Improved |
Closest papers by meaning
- Unsupervised Learning From Video to Detect Foreground Objects in Single Images (ICCV 2017 · similarity 0.97)
- Unsupervised Object Segmentation in Video by Efficient Selection of Highly Probable Positive Features (ICCV 2017 · similarity 0.92)
- Unsupervised Object Localization: Observing the Background To Discover Objects (CVPR 2023 · similarity 0.90)
- Spatially Invariant Unsupervised Object Detection with Convolutional Neural Networks (AAAI 2019 · similarity 0.90)
- Unsupervised Object Detection with Theoretical Guarantees (NeurIPS 2024 · similarity 0.88)
- unMORE: Unsupervised Multi-Object Segmentation via Center-Boundary Reasoning (ICML 2025 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
