Evaluation of Garbage Detection Efficacy Utilising Deep Learning

Early diagnosis of Alzheimer's Disease (AD) is crucial for improving patient quality of life and treatment outcomes. However, accurately classifying MRI scans of AD remains challenging due to the subtle and spatially complex nature of lesion regions. This study proposes a novel bidirectional spatial attention mechanism to enhance the focus on key pathological features in AD MRI images, aiming to improve classification accuracy and support earlier intervention. To enhance model performance, we introduced a customized bidirectional spatial attention module (ATT) integrated into a Swin-Tiny Transformer backbone. Unlike conventional attention methods, the ATT module generates spatial attention maps by adaptively pooling features along both vertical and horizontal orientations, allowing refined adjustment of attention weights across different image regions. Furthermore, to address issues of limited sample size and class imbalance, we employed data augmentation and expansion strategies, enriching the diversity of training data. The model was trained and evaluated on the augmented OASIS1 dataset. The improved Swin-Tiny+ATT model demonstrated significant performance gains across all key metrics on the augmented dataset. Compared to the baseline Swin Transformer, accuracy improved from 84.83% to 87.96%, recall from 89.82% to 91.92%, precision from 85.27% to 91.98%, and the F1 score from 87.26% to 91.89%. These results confirm that the ATT module effectively enhances the model's ability to capture complex spatial features and identify critical lesion regions. The proposed Swin-Tiny+ATT model exhibits strong potential for improving MRI-based classification of Alzheimer's Disease. The bidirectional spatial attention mechanism successfully directs the model's focus to relevant anatomical regions, contributing to higher precision and recall. Combined with data augmentation strategies, the approach mitigates class imbalance and enhances generalization. This work provides a promising deep learning framework to support early and accurate diagnosis of AD, with implications for clinical decision-making and personalized treatment planning.

Garbage classification is a social issue related to people’s livelihood and sustainable development, so letting service robots autonomously perform intelligent garbage classification has important research significance. Aiming at the problems of complex systems with data source and cloud service center data transmission delay and untimely response, at the same time, in order to realize the perception, storage, and analysis of massive multisource heterogeneous data, a garbage detection and classification method based on visual scene understanding is proposed. This method uses knowledge graphs to store and model items in the scene in the form of images, videos, texts, and other multimodal forms. The ESA attention mechanism is added to the backbone network part of the YOLOv5 network, aiming to improve the feature extraction ability of the network, combining with the built multimodal knowledge graph to form the YOLOv5‐Attention‐KG model, and deploying it to the service robot to perform real‐time perception on the items in the scene. Finally, collaborative training is carried out on the cloud server side and deployed to the edge device side to reason and analyze the data in real time. The test results show that, compared with the original YOLOv5 model, the detection and classification accuracy of the proposed model is higher, and the real‐time performance can also meet the actual use requirements. The model proposed in this paper can realize the intelligent decision‐making of garbage classification for big data in the scene in a complex system and has certain conditions for promotion and landing.

With the rise of deep learning technology, intelligent garbage detection provides a new idea for garbage classification management. However, due to the interference of complex environments, coupled with the influence of the irregular features of garbage, garbage detection in complex scenarios still faces significant challenges. Moreover, some of the existing research suffer from shortcomings in either their precision or real-time performance, particularly when applied to complex garbage detection scenarios. Therefore, this paper proposes a model based on YOLOv8, namely HGCS-Det, for detecting garbage in complex scenarios. The HGCS-Det model is designed as follows: Firstly, the normalization attention module is introduced to calibrate the model's attention to targets and to suppress the environmental noise interference information. Additionally, to weigh the attention-feature contributions, an Attention Feature Fusion module is employed to complement the attention weights of each channel. Subsequently, an Instance Boundary Reinforcement module is established to capture the fine-grained features of garbage by combining strong gradient information with semantic information. Finally, the Slide Loss function is applied to dynamically weight hard samples arising from the complex detection environments to improve the recognition accuracy of hard samples. With only a slight increase in parameters (3.02M), HGCS-Det achieves a 93.6% mean average precision (mAP) and 86 FPS on the public HGI30 dataset, which is a 3.33% higher mAP value than from YOLOv12, and outperforms the state-of-the-art (SOTA) methods in both efficiency and applicability. Notably, HGCS-Det maintains a lightweight architecture while enhancing the detection accuracy, enabling real-time performance even in resource-constrained environments. These characteristics significantly improve its practical applicability, making the model well suited for deployment in embedded devices and real-world garbage classification systems. This method can serve as a valuable technical reference for the engineering application of garbage classification.

Garbage detection and classification using a new deep learning-based machine vision system as a tool for sustainable waste recycling

<b><sc>Abstract.</sc></b> In several studies, deep learning algorithms demonstrated feasibility for flower detection and recognition in images. YOLOX is a new high-performance object detector that is the latest in the YOLO (You Only Look Once) deep learning network series that trains on full images and directly optimizes detection performance. No study has been done on peanut flower detection before. This study developed deep learning algorithms using the latest YOLOX models to detect peanut flowers (Arachis hypogaea) from images collected from a peanut research station. YOLOX-L and YOLOX-X models were compared with different configurations to find the best YOLOX object detector for peanut flower detection. Built-in Mosaic and MixUp data augmentation strategies were applied to the training and validation datasets. For experimental purposes, the data augmentation was reduced or enhanced. Another experiment was when the data augmentation strategy, MixUp, was turned off or on. The mean average precision (mAP) evaluation metric was used to evaluate the performance of the YOLOX models. The original 300 epoch recommended by YOLOX developers was tested on the training/validation dataset and was reduced to 100 based on several preliminary tests. YOLOX-X achieved 90.37% mAP with weak/reduced data augmentation on the testing dataset. The average inference time for the unseen images in the testing dataset was 0.134 seconds. YOLOX-X demonstrated feasibility for detecting peanut flowers from in-field acquired images. The presented method will assist researchers in developing a counting method on peanut flowers in images and implementing the detection technique with required minor modifications for other crops or flowers.

One-shot segmentation of novel white matter tracts via extensive data augmentation and adaptive knowledge transfer.

Aiming at the problems of unclear, difficult, and inefficient classification of traditional manual waste, and the difficulty of deploying large existing garbage classification network models, a lightweight garbage detection and classification network S-YOLOv5 is designed based on YOLOv5s. First, a garbage dataset containing 18 types of common household garbage is constructed and labeled according to the principles of garbage classification; secondly, a module combining shufflenetv2 and CoordAttention was introduced to replace the YOLOv5s backbone network, and the ReLU activation function in the shufflenet module was substituted by FReLU; finally the PANet structure was replaced by the BiFPN structure, so as to reduce the model complexity and achieve lightweight while maintaining a high mAP. The experimental results show that the size of S-YOLOv5 is only 2.6MB, which is about 1/6 of the original network size, and the mAP is 80.2%. The size of the proposed network is reduced while maintaining high accuracy, making it more suitable for deployment in smart devices.

With the improvement of people’s living standards, garbage classification is gradually forced. However, due to people’s awareness and knowledge, the classification accuracy and disposal of garbage are difficult to keep pace with guideline changes. With the consideration of the problems of low efficiency, heavy task and poor environment of garbage manual classification, an improved YOLOv7 target detection method is proposed to realize the effective classification of garbage. In this study, the recursive gated convolutional gnconv was used to establish the HorNet network architecture, and the model was trained by making specific data sets. The C3HB module is added to the YOLO model, and the pooling layer is optimized to replace SPPFCSPC to improve the detection accuracy of the target. The experimental results show that the garbage detection and classification method proposed in this study has excellent accuracy. Experiments show that the map value, accuracy and recall rate of the proposed model on garbage datasets are 99.25%, 99.33% and 98.03%, respectively, which are 1.50%, 3.99% and 1.41% higher than those of YOLOv7. The overall results are better than the original model.

Street garbage monitoring is important for ensuring a clean and healthy civic environment. In this context, automatic detection and classification of waste/litter objects on the streets is necessary for proper garbage disposal. Specifically, such detection and classification of waste objects could be utilized in developing automated waste sorting applications. This paper attempts to build a Faster R-CNN based predictive model for automatic classification of ten different types of waste/litter objects. Two pre-trained networks namely Inception-V2 and ResNet-101 are investigated as backbone networks for feature extraction. The performance of the proposed model is also compared with two baselines namely RFCN (region-based fully convolutional network) and SSD (single shot multibox detector). It is observed that the Faster R-CNN configured with InceptionV2 achieves the highest mAP (mean average precision) of 92%.

DAAL-WS: A weakly-supervised method integrated with data augmentation and active learning strategies for MLS point cloud semantic segmentation

Data augmentation is an effective method to prevent model overfitting in deep learning, especially in medical image classification where data samples are small and difficult to obtain. In recent years, different data augmentation methods, such as those based on single data transformation, multiple data mixing, and learning data distribution, have been proposed one after another, but there has never been a systematic system to evaluate various data augmentation methods. An impartial and comprehensive data augmentation evaluation system not only can assess the benefits and drawbacks of existing augmentation approaches in a specific medical image classification but also can provide an effective research direction for the subsequent proposal of new medical image data augmentation methods, thereby advancing the development of auxiliary diagnosis technology based on medical images. Therefore, this paper proposes an objective and universal evaluation system for different data augmentation methods. In this method, different augmented methods are evaluated objectively and comprehensively in terms of classification accuracy and data diversity by using existing large public data sets. The method is universal and easy to operate. To imitate the prevalent small-sized data sets in deep learning, an equal-interval sampling technique based on similarity ranking is presented to select samples from large public data sets and construct a subset that can fully reflect the original set. The augmented data sets are then created using various data augmentation approaches based on the small-sized data sets. Finally, different data augmentation strategies are objectively and fully evaluated based on the comprehensive scores of classification accuracy and data diversity following data augmentation. The validity and feasibility of the suggested sampling method and assessment system in this study are demonstrated by experimental findings on numerous data sets.

The recognition of postearthquake scenes plays an important role in postearthquake rescue and reconstruction. To overcome the over-reliance on expert visual interpretation and the poor recognition performance of traditional machine learning in postearthquake scene recognition, this paper proposes a postearthquake multiple scene recognition (PEMSR) model based on the classical deep learning Single Shot MultiBox Detector (SSD) method. In this paper, a labeled postearthquake scenes dataset is constructed by segmenting acquired remote sensing images, which are classified into six categories: landslide, houses, ruins, trees, clogged and ponding. Due to the insufficiency and imbalance of the original dataset, transfer learning and a data augmentation and balancing strategy are utilized in the PEMSR model. To evaluate the PEMSR model, the evaluation metrics of precision, recall and F1 score are used in the experiment. Multiple experimental test results demonstrate that the PEMSR model shows a stronger performance in postearthquake scene recognition. The PEMSR model improves the detection accuracy of each scene compared with SSD by transfer learning and data augmentation strategy. In addition, the average detection time of the PEMSR model only needs 0.4565s, which is far less than the 8.3472s of the traditional Histogram of Oriented Gradient + Support Vector Machine (HOG+SVM) method.

Garbage classification is an essential work in daily life. With the development of artificial intelligence (AI), we have begun to use object detection to achieve garbage classification. However, the stacking and occlusion of garbage often make the detection performance unsatisfactory. In this paper, YOLOv5 algorithm is used for garbage detection and classification. YOLOv5 has three detection heads, which can detect objects at multiple scales. At the same time, Mosaic and Coarse Dropout image augmentation techniques are used to process our private data set of garbage, which improves the diversity of images. In this way, the target classification accuracy of the detection model can reach 75.7%, which is significantly improved compared with 69.9% without data augmentation. This paper provides a new technical support for the scene classification of stacked garbage based on deep learning.

Search effectiveness for tasks where the retrieval units are clearly defined documents is generally evaluated using standard measures such as mean average precision (MAP). However, many practical speech search tasks focus on content within large spoken files lacking defined structure. These data must be segmented into smaller units for search which may only partially overlap with relevant material. We introduce two new metrics for the evaluation of search effectiveness for informally structured speech data: mean average segment precision (MASP) which measures retrieval performance in terms of both content segmentation and ranking with respect to relevance; and mean average segment distance-weighted precision (MASDWP) which takes into account the distance between the start of the relevant segment and the retrieved segment. We demonstrate the effectiveness of these new metrics on a retrieval test collection based on the AMI meeting corpus.KeywordsSpeech retrievalinformally structured speechevaluation metrics

A supervised data augmentation strategy based on random combinations of key features