The Art of Camouflage: Few-Shot Learning for Animal Detection and Segmentation (2023-04-15T00:00:00.000000Z)

TL;DR

This work proposes FS-CDIS, a framework to efficiently detect and segment camouflaged instances via two loss functions contributing to the training process, and achieves state-of-the-art performance on the newly collected dataset.

Abstract

Camouflaged object detection and segmentation is a new and challenging research topic in computer vision. There is a serious issue of lacking data on concealed objects such as camouflaged animals in natural scenes. In this paper, we address the problem of few-shot learning for camouflaged object detection and segmentation. To this end, we first collect a new dataset, CAMO-FS, for the benchmark. As camouflaged instances are challenging to recognize due to their similarity compared to the surroundings, we guide our models to obtain camouflaged features that highly distinguish the instances from the background. In this work, we propose FS-CDIS, a framework to efficiently detect and segment camouflaged instances via two loss functions contributing to the training process. Firstly, the instance triplet loss with the characteristic of differentiating the anchor, which is the mean of all camouflaged foreground points, and the background points are employed to work at the instance level. Secondly, to consolidate the generalization at the class level, we present instance memory storage with the scope of storing camouflaged features of the same category, allowing the model to capture further class-level information during the learning process. The extensive experiments demonstrated that our proposed method achieves state-of-the-art performance on the newly collected dataset. Code is available at https://github.com/danhntd/FS-CDIS.

Authors

Thanh-Toan Do

5 papers

Tam V. Nguyen

5 papers

M. Tran

4 papers

TL;DR

Abstract

Authors

References95 items

Decoupling Classifier for Boosting Few-shot Object Detection and Instance Segmentation

Deep soft threshold feature separation network for infrared handprint identity recognition and time estimation

Non-linear statistical image watermark detector

Few-Shot Segmentation via Divide-and-Conquer Proxies

Instance-Level Few-Shot Learning With Class Hierarchy Mining

Generating Features with Increased Crop-Related Diversity for Few-Shot Object Detection

Base and Meta: A New Perspective on Few-Shot Segmentation

Reference Twice: A Simple and Unified Baseline for Few-Shot Instance Segmentation

FS-DETR: Few-Shot DEtection TRansformer with prompting and without re-training

Dynamic Transformer for Few-shot Instance Segmentation

Immune Coordination Deep Network for Hand Heat Trace Extraction

Disentangle and Remerge: Interventional Knowledge Distillation for Few-Shot Object Detection from A Conditional Causal Perspective

Holistic Prototype Activation for Few-Shot Segmentation

Kernelized Few-shot Object Detection with Efficient Integral Aggregation

iFS-RCNN: An Incremental Few-shot Instance Segmenter

Matching Feature Sets for Few-Shot Image Classification

Few-Shot Object Detection with Fully Cross-Transformer

Floating pollutant image target extraction algorithm based on immune extremum region

GANORCON: Are Generative Models Useful for Few-shot Segmentation?

Query Adaptive Few-Shot Object Detection with Heterogeneous Graph Convolutional Networks

Partner-Assisted Learning for Few-Shot Image Classification

DeFRCN: Decoupled Faster R-CNN for Few-Shot Object Detection

OpenForensics: Large-Scale Challenging Dataset For Multi-Face Forgery Detection And Segmentation In-The-Wild

Transformation Invariant Few-Shot Object Detection

Inferring Camouflaged Objects by Texture-Aware Interactive Guidance Network

CamouFinder: Finding Camouflaged Instances in Images

Incremental Few-Shot Instance Segmentation

Hallucination Improves Few-Shot Object Detection

Camouflaged Object Segmentation with Distraction Mining

Meta Faster R-CNN: Towards Accurate Few-Shot Object Detection with Attentive Feature Alignment

Uncertainty-aware Joint Salient Object and Camouflaged Object Detection

Mutual Graph Learning for Camouflaged Object Detection

Camouflaged Instance Segmentation In-the-Wild: Dataset, Method, and Benchmark Suite

Dense Relation Distillation with Context-aware Aggregation for Few-Shot Object Detection

Meta-DETR: Image-Level Few-Shot Object Detection with Inter-Class Correlation Exploitation

FSCE: Few-Shot Object Detection via Contrastive Proposal Encoding

Beyond Max-Margin: Class Margin Equilibrium for Few-shot Object Detection

Simultaneously Localize, Segment and Rank the Camouflaged Objects

Semantic Relation Reasoning for Shot-Stable Few-Shot Object Detection

Exploring Complementary Strengths of Invariant and Equivariant Representations for Few-Shot Learning

UniT: Unified Knowledge Transfer for Any-shot Object Detection and Segmentation

Betrayed by Motion: Camouflaged Object Discovery via Motion Segmentation

Generalized Few-shot Semantic Segmentation

MirrorNet: Bio-Inspired Camouflaged Object Segmentation

Few-Shot Object Detection and Viewpoint Estimation for Objects in the Wild

Multi-Scale Positive Sample Refinement for Few-Shot Object Detection

Prototypical Siamese Networks for Few-shot Learning

Self-supervised Knowledge Distillation for Few-shot Learning

Camouflaged Object Detection

FGN: Fully Guided Network for Few-Shot Instance Segmentation

CRNet: Cross-Reference Networks for Few-Shot Segmentation

Frustratingly Simple Few-Shot Object Detection

Meta R-CNN: Towards General Solver for Instance-Level Low-Shot Learning

PANet: Few-Shot Image Semantic Segmentation With Prototype Alignment

Few-Shot Object Detection With Attention-RPN and Multi-Relation Detector

Anabranch network for camouflaged object segmentation

Background matching and disruptive coloration as habitat-specific strategies for camouflage

Few-Shot Object Detection via Feature Reweighting

Superpixel-Enhanced Pairwise Conditional Random Field for Semantic Segmentation

Unsupervised Feature Learning via Non-parametric Instance Discrimination

Dynamic Few-Shot Visual Learning Without Forgetting

Memory Matching Networks for One-Shot Image Recognition

Mask R-CNN

Prototypical Networks for Few-shot Learning

Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks

YOLO9000: Better, Faster, Stronger

Feature Pyramid Networks for Object Detection

Matching Networks for One Shot Learning

It's Moving! A Probabilistic Model for Causal Motion Segmentation in Moving Camera Videos

Deep Residual Learning for Image Recognition

Camouflage performance analysis and evaluation framework based on features fusion

Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks

Foreground object segmentation for moving camera sequences based on foreground-background probabilistic models and prior probability maps

Foreground Object Detection Using Top-Down Information Based on EM Framework

Study on the Camouflaged Target Detection Method Based on 3D Convexity

Superpixel-based object class segmentation using conditional random fields