Self-supervised Robust Object Detectors from Partially Labelled Datasets
- URL: http://arxiv.org/abs/2005.11549v2
- Date: Sat, 27 Jun 2020 03:56:12 GMT
- Title: Self-supervised Robust Object Detectors from Partially Labelled Datasets
- Authors: Mahdieh Abbasi, Denis Laurendeau, Christian Gagne
- Abstract summary: merging datasets allows us to train one integrated object detector, instead of training several ones.
We propose a training framework to overcome missing-label challenge of the merged datasets.
We evaluate our proposed framework for training Yolo on a simulated merged dataset with missing rate $approx!48%$ using VOC2012 and VOC2007.
- Score: 3.1669406516464007
- License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
- Abstract: In the object detection task, merging various datasets from similar contexts
but with different sets of Objects of Interest (OoI) is an inexpensive way (in
terms of labor cost) for crafting a large-scale dataset covering a wide range
of objects. Moreover, merging datasets allows us to train one integrated object
detector, instead of training several ones, which in turn resulting in the
reduction of computational and time costs. However, merging the datasets from
similar contexts causes samples with partial labeling as each constituent
dataset is originally annotated for its own set of OoI and ignores to annotate
those objects that are become interested after merging the datasets. With the
goal of training \emph{one integrated robust object detector with high
generalization performance}, we propose a training framework to overcome
missing-label challenge of the merged datasets. More specifically, we propose a
computationally efficient self-supervised framework to create on-the-fly
pseudo-labels for the unlabeled positive instances in the merged dataset in
order to train the object detector jointly on both ground truth and pseudo
labels. We evaluate our proposed framework for training Yolo on a simulated
merged dataset with missing rate $\approx\!48\%$ using VOC2012 and VOC2007. We
empirically show that generalization performance of Yolo trained on both ground
truth and the pseudo-labels created by our method is on average $4\%$ higher
than the ones trained only with the ground truth labels of the merged dataset.
Related papers
- Anno-incomplete Multi-dataset Detection [67.69438032767613]
We propose a novel problem as "-incomplete Multi-dataset Detection"
We develop an end-to-end multi-task learning architecture which can accurately detect all the object categories with multiple partially annotated datasets.
arXiv Detail & Related papers (2024-08-29T03:58:21Z) - Synthetic data enables faster annotation and robust segmentation for
multi-object grasping in clutter [9.092550803271005]
We propose a synthetic data generation method that minimizes human intervention and makes downstream image segmentation algorithms more robust.
Experiments show that the proposed synthetic scene generation can diminish labelling time dramatically.
Pick-and-place experiments demonstrate that segmentation trained on our hybrid dataset (98.9%, 70%) outperforms the real dataset and a publicly available dataset by (6.7%, 18.8%) and (2.8%, 10%) in terms of labelling and grasping success rate.
arXiv Detail & Related papers (2024-01-24T11:58:30Z) - A Bag-of-Prototypes Representation for Dataset-Level Applications [24.629132557336312]
This work investigates dataset vectorization for two dataset-level tasks: assessing training set suitability and test set difficulty.
We propose a bag-of-prototypes (BoP) dataset representation that extends the image-level bag consisting of patch descriptors to dataset-level bag consisting of semantic prototypes.
BoP consistently shows its advantage over existing representations on a series of benchmarks for two dataset-level tasks.
arXiv Detail & Related papers (2023-03-23T13:33:58Z) - Detection Hub: Unifying Object Detection Datasets via Query Adaptation
on Language Embedding [137.3719377780593]
A new design (named Detection Hub) is dataset-aware and category-aligned.
It mitigates the dataset inconsistency and provides coherent guidance for the detector to learn across multiple datasets.
The categories across datasets are semantically aligned into a unified space by replacing one-hot category representations with word embedding.
arXiv Detail & Related papers (2022-06-07T17:59:44Z) - Primitive3D: 3D Object Dataset Synthesis from Randomly Assembled
Primitives [44.03149443379618]
We propose a cost-effective method for automatically generating a large amount of 3D objects with annotations.
These objects are auto-annotated with part labels originating from primitives.
Considering the large overhead of learning on the generated dataset, we propose a dataset distillation strategy.
arXiv Detail & Related papers (2022-05-25T10:07:07Z) - Learning Semantic Segmentation from Multiple Datasets with Label Shifts [101.24334184653355]
This paper proposes UniSeg, an effective approach to automatically train models across multiple datasets with differing label spaces.
Specifically, we propose two losses that account for conflicting and co-occurring labels to achieve better generalization performance in unseen domains.
arXiv Detail & Related papers (2022-02-28T18:55:19Z) - Semi-supervised Interactive Intent Labeling [6.99674326582747]
We have developed an Intent Bulk Labeling system for SDS developers.
The users can interactively label and augment training data from unlabeled utterance corpora.
We achieve over 10% gain in clustering accuracy on some datasets.
arXiv Detail & Related papers (2021-04-27T18:06:55Z) - Simple multi-dataset detection [83.9604523643406]
We present a simple method for training a unified detector on multiple large-scale datasets.
We show how to automatically integrate dataset-specific outputs into a common semantic taxonomy.
Our approach does not require manual taxonomy reconciliation.
arXiv Detail & Related papers (2021-02-25T18:55:58Z) - Object Detection with a Unified Label Space from Multiple Datasets [94.33205773893151]
Given multiple datasets with different label spaces, the goal of this work is to train a single object detector predicting over the union of all the label spaces.
Consider an object category like faces that is annotated in one dataset, but is not annotated in another dataset.
Some categories, like face here, would thus be considered foreground in one dataset, but background in another.
We propose loss functions that carefully integrate partial but correct annotations with complementary but noisy pseudo labels.
arXiv Detail & Related papers (2020-08-15T00:51:27Z) - Cross-dataset Training for Class Increasing Object Detection [52.34737978720484]
We present a conceptually simple, flexible and general framework for cross-dataset training in object detection.
By cross-dataset training, existing datasets can be utilized to detect the merged object classes with a single model.
While using cross-dataset training, we only need to label the new classes on the new dataset.
arXiv Detail & Related papers (2020-01-14T04:40:47Z)
This list is automatically generated from the titles and abstracts of the papers in this site.
This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.