scispace - formally typeset
Search or ask a question

Answers from top 6 papers

More filters
Papers (6)Insight
Open accessProceedings ArticleDOI
27 Jun 2016
27.2K Citations
Compared to state-of-the-art detection systems, YOLO makes more localization errors but is less likely to predict false positives on background.
This paper presents a robust and efficient ALPR system based on the state-of-the-art YOLO object detector.
Open accessProceedings ArticleDOI
01 Aug 2018
44 Citations
In our experiments, YOLO outperforms the other state-of-the-art detector, Faster R-CNN, and our multi-projection YOLO achieves the best accuracy with low-resolution input.
Proceedings ArticleDOI
01 Oct 2017
62 Citations
A new object detection method, OYOLO (Optimized YOLO), is produced, which is 1.18 times faster than YOLO, while outperforming other region-based approaches like R-CNN in accuracy.
Open accessJournal ArticleDOI
Wei Fang, Wang Lin, Ren Peiming 
01 Jan 2020-IEEE Access
152 Citations
The object detection performance is enhanced in Tinier-YOLO by using the passthrough layer that merges feature maps from the front layers to get fine-grained features, which can counter the negative effect of reducing the model size.
Proceedings ArticleDOI
Zhi Xu, Haochen Shi, Ning Li, Chao Xiang, Huiyu Zhou 
01 Nov 2018
42 Citations
Due to the One-stage detection model like YOLO having novel structure and great industrial application potential, this paper proposes a new model of detection based on YOLOv2 structure.

Related Questions

How to combe YOLO detection with reid for tracking?5 answersTo combine YOLO detection with reid for tracking, an approach is to integrate the attention module CBAM with the Neck part of the YOLOv5s network to improve the target detection model's ability to extract small object features and enhance recognition ability. Additionally, the Market-1501 dataset can be used to train the pedestrian re-id model, and the VOC dataset can be used to train the detection model specifically for detecting pedestrians. Finally, the improved YOLOv5s detector can be connected with DeepSORT, and a virtual detection line can be set in the video to count the flow of people. This approach has shown improved accuracy compared to the original algorithm and achieved a high detection rate in test videos.
How can YOLO detection be improved with a branch for re-identification (ReID) for tracking?4 answersYOLO detection can be improved with a branch for re-identification (ReID) for tracking. One approach is to introduce a lightweight CARAFE module for up-sampling on the neck layer, which reduces complexity and improves the learning ability of the network. Another method is to design a transformer encoder layer (TE-Layer) and a convolutional block attention module (CBAM) to capture global and rich contextual information, and implement cross-domain object detection through knowledge distillation. Additionally, a context enhancement module can be used to capture global and local information of the image, and high- and low-frequency information decomposed by wavelet transform can be introduced for multi-scale feature fusion. These improvements enhance the sensitivity and detection performance of YOLOv5, making it better suited for tracking and re-identification tasks.
What is accuracy in yolo to detect object?5 answersAccuracy in YOLO for object detection refers to the ability of the algorithm to correctly identify and locate objects in an image. Several papers have discussed the accuracy of different versions of YOLO. Safaldin et al. proposed an improved YOLO-v2 for detecting tiny objects and achieved a detection accuracy of 95.8%. Another paper introduced ACP-YOLO, an enhanced version of YOLOv4, which achieved an average precision of 45.6% and surpassed YOLOv4 in terms of accuracy. Additionally, a paper on SuperYOLO demonstrated its accuracy in small object detection, achieving an accuracy of 73.61% on the VEDAI RS dataset. These papers highlight the advancements made in improving the accuracy of YOLO for object detection.
How can we make license plate recognition using YOLO more efficient?5 answersTo make license plate recognition using YOLO more efficient, several methods have been proposed in the literature. One approach is to add an improved channel attention mechanism to the down-sampling process of YOLOv5, which enhances the feature extraction ability of the model. Another method is to reduce the number of parameters on the input side and set only one class in the YOLO layer, which improves the efficiency and accuracy of the detector for locating license plates. Additionally, using Gated recurrent units (GRU) + Connectionist temporal classification (CTC) for character segmentation-free recognition can significantly shorten the training time and improve the convergence speed and recognition accuracy of the network. Another approach is to propose an end-to-end license plate optimization recognition algorithm based on YOLOv3, which includes a detection dewarping convolutional neural network (DU-CNN) to improve the recognition ability of license plates. These methods have shown better recognition accuracy, improved efficiency, and stability in license plate recognition tasks.
Why YOLO-based road monitoring systems help in real-time vehicle detection?5 answersYOLO-based road monitoring systems are beneficial for real-time vehicle detection due to their ability to track, identify, and categorize vehicles, count the number of vehicles, and classify them accurately and efficiently. The YOLO architecture, combined with OpenCV, enhances vehicle detection accuracy and execution time, even in challenging traffic patterns, through techniques such as transfer learning and extensive dataset collection. Additionally, the use of YOLO and OpenCV trackers, such as CSRT and KCF, has shown superior performance in vehicle detection, tracking, and counting accuracy, achieving high percentages of accuracy in both real-time traffic video datasets and specific KNUVDB datasets. Furthermore, the proposed MEB-YOLO model, incorporating data augmentation, attention mechanisms, and feature fusion, has demonstrated improved detection results on public datasets and real traffic monitoring data, making it efficient for detecting road targets.
What is YOLO?2 answersYOLO (You Only Look Once) is a popular algorithm in machine learning that is used for target detection. It has rapidly developed with several versions available. Each version has its own unique features and application areas. YOLO is known for its fast detection speed and high accuracy compared to previous algorithms. It is widely used in various sectors such as autonomous driving, video surveillance, face recognition, and more. YOLO models are designed to detect different objects under different circumstances, making them versatile. The algorithm has been applied in accident detection systems, crop pathology identification, and industrial surface defect detection. YOLO's evolution has focused on real-time performance and high classification accuracy, making it suitable for deployment on constrained edge devices.

See what other people are reading

What are the factors that contribute to the resilience of the human brain?
3 answers
Resilience of the human brain is influenced by a combination of genetic and environmental factors. Lifestyle factors such as sleep quality, physical activity, diet, smoking, alcohol consumption, social contact, and education play a role in promoting structural brain resilience. Nature, or genetics, is a stronger factor in resilience, but the contribution of nurture, or the environment, should also be considered. Glycolytic metabolism, the breakdown of glucose, and glycolytic dysfunction have been linked to brain aging and neurodegenerative diseases, including Alzheimer's disease (AD). Neuronal glycolysis is affected by the ApoE gene isoforms, and glycolytic robustness may render the brain more resistant to neurodegenerative risks. Brain resilience is influenced by genetic factors related to region-to-region communication in functional brain networks, particularly the Default Mode, Visual, and Sensorimotor networks. These factors contribute to the resilience of the human brain and may inform interventions to mitigate accelerated aging and reduce the risk of neurodegenerative diseases.
How can teachers' English proficiency be improved?
4 answers
Teachers' English proficiency can be improved by implementing various strategies. One approach is to establish a teaching quality evaluation system (TQES) that allows teachers to receive feedback and identify areas for improvement in their teaching process. Another effective method is the use of a shadowing video-based program, which has been found to significantly enhance teaching skills and English language proficiency. Additionally, teachers can help students by making them aware of the importance of their contribution to language learning and teaching them language learning strategies. Furthermore, a reconceptualization of teacher language proficiency as a specialized subset of language skills required for teaching can be beneficial. This concept of English-for-Teaching focuses on specific classroom language and can enhance teachers' ability to manage the classroom, communicate lesson content, and assess students.
What is Machine learning?
5 answers
Machine learning is a branch of artificial intelligence that involves teaching computers to learn from examples, data, and experience without being explicitly programmed. It uses algorithms to enable computers to evolve behaviors based on empirical data. Machine learning is closely related to other fields such as statistics, computer science, and artificial intelligence. It has been successfully applied in various fields including pattern recognition, computer vision, finance, and biomedical and medical applications. Machine learning algorithms have the potential to optimize and automate complex processes, improve decision-making, and enhance outcomes in fields such as radiotherapy. Machine learning addresses problems where there are no human experts, problems where human experts are unable to explain their expertise, problems with rapidly changing phenomena, and applications that need to be customized for individual users.
What are extrinsic and intrinsic motivation?
3 answers
Extrinsic motivation refers to the drive to perform a behavior in order to obtain external rewards or avoid punishment. It is influenced by factors such as pay, compensation, and communication within a team. Intrinsic motivation, on the other hand, is the internal desire to engage in a behavior for its own sake, driven by personal interest, enjoyment, or satisfaction. It is not dependent on external rewards or punishments. Intrinsic motivation can be further classified into different types, such as intrinsic motivation to know, intrinsic motivation toward accomplishments, and intrinsic motivation to experience stimulation. Understanding the distinction between extrinsic and intrinsic motivation is important in various contexts, including team performance, reinforcement learning tasks, psychiatric illness, and foreign language learning.
How does aes encryption work?
5 answers
The AES encryption algorithm works by dividing the message into blocks of 128, 192, or 256 bits. It then applies a series of calculations, including the ShiftRows process and three substitution processes: SubBytes, MixColumns, and AddRoundKey. These calculations are designed to provide a high level of security and confidentiality. The encryption procedure is efficient and has a high level of security. The S-box module in AES plays a crucial role in providing confusion and diffusion measures during encryption. To enhance the throughput rate and optimize path delay, composite field arithmetic-based Sub-bytes and inverse Sub-bytes operations are used. The AES algorithm is widely used and offers a high level of security. It provides different key sizes and block sizes, and understanding the critical aspects of the algorithm is essential for successful encryption and decryption. Researchers have also designed hardware implementations of the AES algorithm using VHDL, resulting in faster encryption and decryption processes.
Nueropathy related research paper due to diabetes based on Imaging Studies: and genetic data?
5 answers
Diabetes is a metabolic disorder that can lead to various complications, including neuropathy. Imaging studies and genetic data have been used to understand the relationship between diabetes and neuropathy. Multiple neuroimaging measures have been found to be heritable and may reflect the pathogenesis of cognitive decline and dementia in patients with type 2 diabetes. However, no significant associations were found between candidate single-nucleotide polymorphisms (SNPs) and neuroimaging phenotypes in prior studies. A missense variant in the PLEKHG4B gene was significantly associated with white matter and gray matter mean diffusivity, suggesting a genetic contribution to variation in neuroimaging measures. Deep learning methods have also been proposed for the early detection of diabetes, using numerical data converted into images for classification. Overall, these studies highlight the importance of both imaging studies and genetic data in understanding and detecting neuropathy related to diabetes.
Why to use Fuzzy Mamdani Clustering Implementation?
3 answers
Fuzzy Mamdani clustering implementation is used because it can handle phenomena that are ambiguous, disguised, or fuzzy. It is a mathematical framework that represents uncertainty, ambiguity, imprecision, lack of information, and partial truth. The Mamdani method is a way to map an input space into an output space using fuzzy logic. It can be used to calculate values based on various criteria for evaluating teacher performance. The Mamdani method is also used in the selection process of new students at a tertiary institution based on predetermined criteria. Additionally, it is used in designing a fuzzy logic system for diagnosing diabetes in human beings. The Mamdani method has the ability to process data based on determining criteria, making it suitable for these applications.
What are the ontological underpinnings of qualitative research?
5 answers
Qualitative research is underpinned by various ontological perspectives. Researchers recognize that realities are multiple and subjective, and they aim to understand phenomena from the perspectives of participants. The ontological stance of qualitative research is in contrast to the positivist, quantitative approach that assumes an objective reality. The researcher's ontological stance influences every aspect of the research process, including the chosen methodology, methods, research questions, and prospective outcomes. Researchers engaging in qualitative studies online must closely examine and interrogate the digital tools, systems, and services they use to establish trustworthy epistemological claims. The philosophical assumptions underlying qualitative research are integral to the procedures and frameworks employed, such as phenomenology, which sets the stage for discussion and process illustration.
How can AI help in the fight against cybercrime?
5 answers
AI can help in the fight against cybercrime by enabling more advanced and efficient threat detection and response. AI-powered systems can analyze vast amounts of data and identify patterns that would be difficult or impossible for a human to detect, allowing for real-time threat response. Additionally, AI can help organizations better manage and secure their networks and devices, as well as identify and mitigate vulnerabilities. AI-based solutions can provide effective and robust cyber defense capabilities, including identifying malware attacks, network intrusions, phishing and spam emails, and data breaches, and alerting security incidents when they occur. However, it is important to note that AI also presents new challenges in cybersecurity, as it can be used to enable more sophisticated forms of cyber-attacks.
Concept of psychological test for counseling?
4 answers
Psychological tests are used in counseling to assess and understand the psychological state of individuals. These tests involve the use of various features such as data variance weighted information entropy, gradient direction features, and local contrast features. The complexity of the image is defined by combining these features, which allows for effective feature extraction and segmentation in psychological counseling. The choice of appropriate tests and the proper interpretation of test scores are crucial in counseling. Counselors need to have statistical and technical knowledge about tests, as well as substantive psychological knowledge about the behavior domain being assessed. Psychological assessment is considered a core competency in counseling, providing specific answers to various clinical, psychoeducational, occupational, forensic, and neuropsychological questions. Best practices in assessment include attention to diversity contexts, ethics, and standards.
What are the most popular features of TikTok?
4 answers
TikTok's most popular features include the ability for content creators to use stylistic devices such as semantic and lexical parallelism to emphasize specific ideas and create humorous effects. Additionally, TikTok provides a platform for women to articulate their interpretation of their bodies, freeing themselves from gendered prescriptions and becoming the subject of their bodies. The app is particularly popular among young people, with teenagers being the main user demographic. TikTok is also known for its memetic videos, which often feature lip-syncing, dance routines, and comedic skits. Furthermore, TikTok has become a theatrical space where users can engage in performance-making, blurring the lines between live performance, theatricality, and digital culture.