Inicio  /  Applied Sciences  /  Vol: 13 Par: 18 (2023)  /  Artículo
ARTÍCULO
TITULO

Applying Object Detection and Embedding Techniques to One-Shot Class-Incremental Multi-Label Image Classification

Youngki Park and Youhyun Shin    

Resumen

In this paper, we introduce an efficient approach to multi-label image classification that is particularly suited for scenarios requiring rapid adaptation to new classes with minimal training data. Unlike conventional methods that rely solely on neural networks trained on known classes, our model integrates object detection and embedding techniques to allow for the fast and accurate classification of novel classes based on as few as one sample image. During training, we use either Convolutional Neural Network (CNN)- or Vision Transformer-based algorithms to convert the provided sample images of new classes into feature vectors. At inference, a multi-object image is analyzed using low-threshold object detection algorithms, such as YOLOS or CutLER, identifying virtually all object-containing regions. These regions are subsequently converted into candidate vectors using embedding techniques. The k-nearest neighbors are identified for each candidate vector, and labels are assigned accordingly. Our empirical evaluation, using custom multi-label datasets featuring random objects and backgrounds, reveals that our approach substantially outperforms traditional methods lacking object detection. Notably, unsupervised object detection exhibited higher speed and accuracy than its supervised counterpart. Furthermore, lightweight CNN-based embeddings were found to be both faster and more accurate than Vision Transformer-based methods. Our approach holds significant promise for applications where classes are either rarely represented or continuously evolving.

 Artículos similares

       
 
Feihu Zhang, Wei Zhang, Chensheng Cheng, Xujia Hou and Chun Cao    
Deep learning-based object detection methods have demonstrated remarkable effectiveness across various domains. Recently, there has been growing interest in applying these techniques to underwater environments. Conventional optical imaging methods face s... ver más

 
Xuan Viet Linh Nguyen, Tony Gerges, Pascal Bevilacqua, Jean-Marc Duchamp, Philippe Benech, Jacques Verdier, Philippe Lombard, Pangsui Usifu Linge, Fabien Mieyeville, Michel Cabrera and Bruno Allard    
Harvesting of ambient radio-frequency energy is largely covered in the literature. The RF energy harvester is considered most of the time as a standalone board. There is an interest to add the RF harvesting function on an already-designed object. Polymer... ver más

 
Yifan Zhang, Guoyou Shi and Jiao Liu    
The unmanned surface vehicle (USV) is significantly affected by the ocean environment and weather conditions when navigating. The energy consumption is large, which is not conducive to completing water tasks. This study investigates the global energy-eff... ver más

 
Tomislav Pe?a, Maja Krcum, Grgo Kero and Jo?ko ?oda    
The ship?s power system is one of the most important systems on board. It is designed for uninterrupted power supply to all ship consumers under different conditions of exploitation. When designing a ship, various optimizations are conducted to build the... ver más

 
Shih-Chang Huang and Cong-Han Huang    
Detecting the intangible continuous object (ICO) is a significant task, especially when the ICO is harmful as a toxic gas. Many studies used steady sensors to sketch the contour and find the area of the ICO. Applying the mobile sensors can further improv... ver más
Revista: Algorithms