Confidence-Aware Object Detection Based on MobileNetv2 for Autonomous Driving

Li, Wei; Li, Kai

doi:10.3390/s21072380

Cited by 14 publications

(7 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…DeepLabV3+ is a semantic segmentation model that combines Encoder and Decoder components [31]. The Encoder utilizes the Xception backbone network and deep features extracted by Atrous Spatial Pyramid Pooling (ASPP) [32]. These features are then fed into the Decoder through upsampling, where they are fused with the original shallow feature map.…”

Section: Identification Of Field and Road Boundariesmentioning

confidence: 99%

See 1 more Smart Citation

Design and Test of Intelligent Farm Machinery Operation Control Platform for Unmanned Farms

Wang,

Yue,

Yang

et al. 2024

Agronomy

View full text Add to dashboard Cite

The unmanned farm control platform is of great significance in promoting the supervision of farm production with less manpower or autonomous operation of farm machinery and the construction of farm informatization. Addressing the existing control platform for farm location information acquisition is time-consuming, labor-intensive, and lacks the whole process control of multiple types of farm machinery. In this paper, we propose an Internet of Things (IoT) control scheme for intelligent farm machinery operation of unmanned farms and design the access standards for multiple types of farm machinery, as well as realize the remote control of intelligent farm machinery operation by constructing a remote control model. A high-precision map construction method is designed to improve the DeepLabV3+ algorithm to identify fields and roads. The control models of path planning, remote task, remote control, and safety system are built to achieve the remote control of intelligent agricultural machinery operation. The proposed technology is implemented in the platform integration and application tests are carried out. The error of the constructed high-precision map is less than 3 cm, the completeness rate of the automatic boundary extraction rate is 96.71%, and the correctness rate is 95.63%, which can be used to obtain the boundary instead of manual labeling or on-site point picking. The use of the platform for the simultaneous control of three farm machinery operations reduces the number of people in operation and production and reduces the professional requirements of the personnel, which will promote the management of the entire farm by one person or even by no one in the future.

show abstract

Section: Identification Of Field and Road Boundariesmentioning

confidence: 99%

“…features extracted by Atrous Spatial Pyramid Pooling (ASPP) [32]. These features are then fed into the Decoder through upsampling, where they are fused with the original shallow feature map.…”

Section: Identification Of Field and Road Boundariesmentioning

confidence: 99%

Design and Test of Intelligent Farm Machinery Operation Control Platform for Unmanned Farms

Wang,

Yue,

Yang

et al. 2024

Agronomy

View full text Add to dashboard Cite

show abstract

“…MobileNetV2 utilizes an inverted residual structure and linear bottleneck to create an efficient layer structure. It employs separable depth-wise convolutions, which split the standard convolution into two separate functions [30]. This approach reduces computational requirements and the overall model size, as depicted in Fig.…”

Section: ) Mobilenetv2 Modelmentioning

confidence: 99%

Deep Learning-Based Lane-Keeping Assist System for Self-Driving Cars Using Transfer Learning and Fine Tuning

Hong,

Khanh,

Vinh

et al. 2024

JAIT

View full text Add to dashboard Cite

This paper presents an advanced lane-keeping assistance system specifically designed for self-driving cars. The proposed model combines the powerful Xception network with transfer learning and fine-tuning techniques to accurately predict the steering angle. By analyzing cameracaptured images, the model effectively learns from human driving knowledge and provides precise estimations of the steering angle necessary for safe lane-keeping. The transfer learning technique allows the model to leverage the extensive knowledge acquired from the ImageNet dataset, while the fine-tuning technique is utilized to tailor the pre-trained model to the specific task of steering angle prediction based on input images, enabling optimal performance. Fine-tuning was initiated by initially freezing the pre-trained model and training only the Fully Connected (FC) layer for the first 10 epochs. Subsequently, the entire model, encompassing both the backbone and the FC layer, was unfrozen for further training. To evaluate the system's effectiveness, a comprehensive comparative analysis is conducted against popular existing models, including Nvidia, MobilenetV2, VGG19, and InceptionV3. The evaluation includes an assessment of the operational accuracy based on the loss function, specifically utilizing the Mean Squared Error (MSE) equation. The proposed model achieves the lowest loss function values for both training and validation, demonstrating its superior predictive performance. Additionally, the model's performance is further evaluated through extensive real-world testing on pre-designed trajectories and maps, resulting in the minimal deviation of the steering angle from the desired trajectory over time. This practical evaluation provides valuable insights into the mode's reliability and its potential to effectively assist in lanekeeping tasks.

show abstract

“…The coverage area of the HAP is large enough to ensure that GVs are always and continuously under coverage in the AoI, even in case of mobility Each GV generates video frames of size n UL ∈ [1, 3] Mb from its camera sensor at an average rate r = 10 fps. Object detection on these frames requires a constant computational load of C = 60 GFLOP per frame, which is computed as the average between the computational performance of two popular object detectors, namely Gaussian YOLO and SqueezeDet+ [18]. If frames are offloaded to the HAP (with probability η * ), eventually the processed output (i.e., the bounding boxes of the detected objects) is returned to the GVs in a packet of a much smaller size than the original frame, i.e., n DL = 100 kb, which implies that t DL ≪ t UL .…”

Section: A Simulation Setup and Parametersmentioning

confidence: 99%

Real-Time HAP-Assisted Vehicular Edge Computing for Rural Areas

Traspadini

Giordani

Giambene

et al. 2023

IEEE Wireless Commun. Lett.

View full text Add to dashboard Cite

Non-Terrestrial Networks (NTNs) are expected to be a key component of 6th generation (6G) networks to support broadband seamless Internet connectivity and expand the coverage even in rural and remote areas. In this context, High Altitude Platforms (HAPs) can act as edge servers to process computational tasks offloaded by energy-constrained terrestrial devices such as Internet of Things (IoT) sensors and ground vehicles (GVs). In this paper, we analyze the opportunity to support Vehicular Edge Computing (VEC) via HAP in a rural scenario where GVs can decide whether to process data onboard or offload them to a HAP. We characterize the system as a set of queues in which computational tasks arrive according to a Poisson arrival process. Then, we assess the optimal VEC offloading factor to maximize the probability of real-time service, given latency and computational capacity constraints.

show abstract

Confidence-Aware Object Detection Based on MobileNetv2 for Autonomous Driving

Cited by 14 publications

References 20 publications

Design and Test of Intelligent Farm Machinery Operation Control Platform for Unmanned Farms

Design and Test of Intelligent Farm Machinery Operation Control Platform for Unmanned Farms

Deep Learning-Based Lane-Keeping Assist System for Self-Driving Cars Using Transfer Learning and Fine Tuning

Real-Time HAP-Assisted Vehicular Edge Computing for Rural Areas

Contact Info

Product

Resources

About