Towards Open World Object Detection

@article{Joseph2021TowardsOW,
  title={Towards Open World Object Detection},
  author={K. J. Joseph and Salman Hameed Khan and Fahad Shahbaz Khan and Vineeth N. Balasubramanian},
  journal={2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
  year={2021},
  pages={5826-5836}
}
Humans have a natural instinct to identify unknown object instances in their environments. The intrinsic curiosity about these unknown instances aids in learning about them, when the corresponding knowledge is eventually available. This motivates us to propose a novel computer vision problem called: ‘Open World Object Detection’, where a model is tasked to: 1) identify objects that have not been introduced to it as ‘unknown’, without explicit supervision to do so, and 2) incrementally learn… 
Expanding Low-Density Latent Regions for Open-Set Object Detection
TLDR
This work proposes to identify unknown objects by separating high/low-density regions in the latent space, based on the consensus that unknown objects are usually distributed in low-density latent regions, and presents a novel Open-set Detector (OpenDet) with expanded low- density regions.
Towards Open-Set Object Detection and Discovery
TLDR
This work proposes a two-stage method that uses an open-set object detector to predict both known and unknown objects and study the representation of predicted objects in an unsupervised manner and discover new categories from the set of unknown objects.
VOS: Learning What You Don't Know by Virtual Outlier Synthesis
TLDR
VOS is presented, a novel framework for OOD detection by adaptively synthesizing virtual outliers that can meaningfully regularize the model’s decision boundary during training, which achieves stateof-the-art performance on both object detection and image classification models.
Objects in Semantic Topology
TLDR
Semantic topology, either randomly-generated or derived from a well-trained language model, could outperform the current state-of-the-art open-world object detectors by a large margin, e.g., the absolute open-set error is reduced from 7832 to 2546, exhibiting the inherent superiority of semantic topology on open- world object detection.
CODA: A Real-World Road Corner Case Dataset for Object Detection in Autonomous Driving
TLDR
This work introduces a challenging dataset named CODA, which consists of 1500 carefully selected real-world driving scenes, each containing four object-level corner cases (on average), spanning 30+ object categories, and investigates the performance of standard object detectors trained on large-scale autonomous driving datasets on corner cases.
Continual Object Detection via Prototypical Task Correlation Guided Gating Mechanism
TLDR
A unified framework is shared by all tasks while task-aware gates are introduced to automatically select sub-models for specific tasks and various knowledge can be successively mem-orized by storing their corresponding sub-model weights in this system.
Revisiting Open World Object Detection
TLDR
This paper proposes five fundamental benchmark principles to guide the OWOD benchmark construction and introduces a novel and effective OWOD framework containing an auxiliary Proposal ADvisor and a Class-specific Expelling Classifier (CEC).
Class-agnostic Object Detection with Multi-modal Transformer
TLDR
This paper advocates that existing methods lack a top-down supervision signal governed by human-understandable semantics and demonstrates that Multi-modal Vision Transformers (MViT) trained with aligned image-text pairs can effec-tively bridge this gap.
Multi-modal Transformers Excel at Class-agnostic Object Detection
TLDR
This paper advocates that existing methods lack a top-down supervision signal governed by human-understandable semantics and develops an efficient and flexible MViT architecture using multi-scale feature processing and deformable self-attention that can adaptively generate proposals given a specific language query.
ORDER: Open World Object Detection on Road Scenes
TLDR
This work proposes Open World Object Detection on Road Scenes (ORDER) and introduces Feature-Mix to improve the unknown object detection capabilities of an object detector, and proposes a novel loss: Focal regression loss that collectively addresses the problem of small object detection and intra-class bounding box scale variations.
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 65 REFERENCES
The Pascal Visual Object Classes (VOC) Challenge
TLDR
The state-of-the-art in evaluated methods for both classification and detection are reviewed, whether the methods are statistically different, what they are learning from the images, and what the methods find easy or confuse.
The Overlooked Elephant of Object Detection: Open Set
TLDR
This paper is the first to formalize the problem of open-set object detection and propose the first open- set object detection protocol and provides a new evaluation metric to analyze the performance of some state-of-the-art detectors and discusses their performance differences.
Dropout Sampling for Robust Object Detection in Open-Set Conditions
TLDR
It is demonstrated how label uncertainty can be extracted from a state-of-the-art object detection system via Dropout Sampling and used to increase object detection performance under the open-set conditions that are typically encountered in robotic vision.
Energy-based Out-of-distribution Detection
TLDR
This work proposes a unified framework for OOD detection that uses an energy score, and shows that energy scores better distinguish in- and out-of-distribution samples than the traditional approach using the softmax scores.
Faster ILOD: Incremental Learning for Object Detectors based on Faster RCNN
Focal Loss for Dense Object Detection
TLDR
This paper proposes to address the extreme foreground-background class imbalance encountered during training of dense detectors by reshaping the standard cross entropy loss such that it down-weights the loss assigned to well-classified examples, and develops a novel Focal Loss, which focuses training on a sparse set of hard examples and prevents the vast number of easy negatives from overwhelming the detector during training.
Frustratingly Simple Few-Shot Object Detection
TLDR
This work finds that fine-tuning only the last layer of existing detectors on rare classes is crucial to the few-shot object detection task, and establishes a new state of the art on the revised benchmarks.
GDumb: A Simple Approach that Questions Our Progress in Continual Learning
We discuss a general formulation for the Continual Learning (CL) problem for classification—a learning task where a stream provides samples to a learner and the goal of the learner, depending on the
Generative-Discriminative Feature Representations for Open-Set Recognition
TLDR
This work proposes two techniques to force class activations of open-set samples to be low, and uses self-supervision to force the network to learn more informative featues when assigning class scores to improve separation of classes from each other and from open- set samples.
Mask R-CNN
TLDR
This work presents a conceptually simple, flexible, and general framework for object instance segmentation that outperforms all existing, single-model entries on every task, including the COCO 2016 challenge winners.
...
1
2
3
4
5
...