Transfer learning with generative models for object detection on limited datasets

• URL: http://arxiv.org/abs/2402.06784v2
• Date: Thu, 13 Jun 2024 10:09:51 GMT
• Title: Transfer learning with generative models for object detection on limited datasets
• Authors: Matteo Paiano, Stefano Martina, Carlotta Giannelli, Filippo Caruso,
• Abstract summary: In some fields, such as marine biology, it is necessary to have correctly labeled bounding boxes around each object. We propose a transfer learning framework that is valid for a generic scenario. Our results pave the way for new generative AI-based protocols for machine learning applications in various domains.
• Score: 1.4999444543328293
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: The availability of data is limited in some fields, especially for object detection tasks, where it is necessary to have correctly labeled bounding boxes around each object. A notable example of such data scarcity is found in the domain of marine biology, where it is useful to develop methods to automatically detect submarine species for environmental monitoring. To address this data limitation, the state-of-the-art machine learning strategies employ two main approaches. The first involves pretraining models on existing datasets before generalizing to the specific domain of interest. The second strategy is to create synthetic datasets specifically tailored to the target domain using methods like copy-paste techniques or ad-hoc simulators. The first strategy often faces a significant domain shift, while the second demands custom solutions crafted for the specific task. In response to these challenges, here we propose a transfer learning framework that is valid for a generic scenario. In this framework, generated images help to improve the performances of an object detector in a few-real data regime. This is achieved through a diffusion-based generative model that was pretrained on large generic datasets. With respect to the state-of-the-art, we find that it is not necessary to fine tune the generative model on the specific domain of interest. We believe that this is an important advance because it mitigates the labor-intensive task of manual labeling the images in object detection tasks. We validate our approach focusing on fishes in an underwater environment, and on the more common domain of cars in an urban setting. Our method achieves detection performance comparable to models trained on thousands of images, using only a few hundreds of input data. Our results pave the way for new generative AI-based protocols for machine learning applications in various domains.

Related papers

• Generalized Few-Shot 3D Object Detection of LiDAR Point Cloud for Autonomous Driving [91.39625612027386]
  We propose a novel task, called generalized few-shot 3D object detection, where we have a large amount of training data for common (base) objects, but only a few data for rare (novel) classes. Specifically, we analyze in-depth differences between images and point clouds, and then present a practical principle for the few-shot setting in the 3D LiDAR dataset. To solve this task, we propose an incremental fine-tuning method to extend existing 3D detection models to recognize both common and rare objects.
  arXiv  Detail & Related papers  (2023-02-08T07:11:36Z)
• A Multi Camera Unsupervised Domain Adaptation Pipeline for Object Detection in Cultural Sites through Adversarial Learning and Self-Training [23.186208885878926]
  We present a new dataset collected in a cultural site to study the problem of domain adaptation for object detection. We present a new domain adaptation method which outperforms current state-of-the-art approaches.
  arXiv  Detail & Related papers  (2022-10-03T10:40:58Z)
• Neural-Sim: Learning to Generate Training Data with NeRF [31.81496344354997]
  We present the first fully differentiable synthetic data pipeline that uses Neural Radiance Fields (NeRFs) in a closed-loop with a target application's loss function. Our approach generates data on-demand, with no human labor, to maximize accuracy for a target task.
  arXiv  Detail & Related papers  (2022-07-22T22:48:33Z)
• Learning from Temporal Spatial Cubism for Cross-Dataset Skeleton-based Action Recognition [88.34182299496074]
  Action labels are only available on a source dataset, but unavailable on a target dataset in the training stage. We utilize a self-supervision scheme to reduce the domain shift between two skeleton-based action datasets. By segmenting and permuting temporal segments or human body parts, we design two self-supervised learning classification tasks.
  arXiv  Detail & Related papers  (2022-07-17T07:05:39Z)
• Unsupervised Domain Adaptive Learning via Synthetic Data for Person Re-identification [101.1886788396803]
  Person re-identification (re-ID) has gained more and more attention due to its widespread applications in video surveillance. Unfortunately, the mainstream deep learning methods still need a large quantity of labeled data to train models. In this paper, we develop a data collector to automatically generate synthetic re-ID samples in a computer game, and construct a data labeler to simultaneously annotate them.
  arXiv  Detail & Related papers  (2021-09-12T15:51:41Z)
• Unsupervised Domain Adaption of Object Detectors: A Survey [87.08473838767235]
  Recent advances in deep learning have led to the development of accurate and efficient models for various computer vision applications. Learning highly accurate models relies on the availability of datasets with a large number of annotated images. Due to this, model performance drops drastically when evaluated on label-scarce datasets having visually distinct images.
  arXiv  Detail & Related papers  (2021-05-27T23:34:06Z)
• Streaming Self-Training via Domain-Agnostic Unlabeled Images [62.57647373581592]
  We present streaming self-training (SST) that aims to democratize the process of learning visual recognition models. Key to SST are two crucial observations: (1) domain-agnostic unlabeled images enable us to learn better models with a few labeled examples without any additional knowledge or supervision; and (2) learning is a continuous process and can be done by constructing a schedule of learning updates.
  arXiv  Detail & Related papers  (2021-04-07T17:58:39Z)
• Six-channel Image Representation for Cross-domain Object Detection [17.854940064699985]
  Deep learning models are data-driven and the excellent performance is highly dependent on the abundant and diverse datasets. Some image-to-image translation techniques are employed to generate some fake data of some specific scenes to train the models. We propose to inspire the original 3-channel images and their corresponding GAN-generated fake images to form 6-channel representations of the dataset.
  arXiv  Detail & Related papers  (2021-01-03T04:50:03Z)
• A Review of Single-Source Deep Unsupervised Visual Domain Adaptation [81.07994783143533]
  Large-scale labeled training datasets have enabled deep neural networks to excel across a wide range of benchmark vision tasks. In many applications, it is prohibitively expensive and time-consuming to obtain large quantities of labeled data. To cope with limited labeled training data, many have attempted to directly apply models trained on a large-scale labeled source domain to another sparsely labeled or unlabeled target domain.
  arXiv  Detail & Related papers  (2020-09-01T00:06:50Z)
• Co-training for On-board Deep Object Detection [0.0]
  Best performing deep vision-based object detectors are trained in a supervised manner by relying on human-labeled bounding boxes. Co-training is a semi-supervised learning method for self-labeling objects in unlabeled images. We show how co-training is a paradigm worth to pursue for alleviating object labeling, working both alone and together with task-agnostic domain adaptation.
  arXiv  Detail & Related papers  (2020-08-12T19:08:59Z)
• Learning Cross-domain Generalizable Features by Representation Disentanglement [11.74643883335152]
  Deep learning models exhibit limited generalizability across different domains. We propose Mutual-Information-based Disentangled Neural Networks (MIDNet) to extract generalizable features that enable transferring knowledge to unseen categorical features in target domains. We demonstrate our method on handwritten digits datasets and a fetal ultrasound dataset for image classification tasks.
  arXiv  Detail & Related papers  (2020-02-29T17:53:16Z)

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.