convolutional neural network
PulseAugur coverage of convolutional neural network — every cluster mentioning convolutional neural network across labs, papers, and developer communities, ranked by signal.
- instance of alphaXiv 90%
- instance of Vgg16 90%
- instance of MobileNetV2 90%
- instance of ResNet18 90%
- competes with Vision Transformers 70%
- instance of Vision Transformers 70%
- instance of ScienceCast 70%
- instance of Gotit.pub 70%
- used by Vision Transformers 70%
- instance of CatalyzeX 70%
- used by alphaXiv 70%
- used by CatalyzeX 70%
20 day(s) with sentiment data
-
New iBKD framework transfers CNN inductive biases to Vision Transformers under data scarcity
Researchers have developed a new knowledge distillation framework called iBKD, designed to improve the performance of Vision Transformers (ViTs) when training data is limited. This method effectively transfers the induc…
-
New CNN improves bearing fault diagnosis in noisy conditions
Researchers have developed a novel convolutional neural network designed for bearing fault diagnosis, particularly effective in noisy environments. This network utilizes a dual-domain approach, incorporating both time-d…
-
Deep learning framework accurately detects repetitive behaviors using wearable sensors
Researchers have developed a deep learning framework using multimodal wearable sensor data to accurately detect and classify body-focused repetitive behaviors like hair pulling and skin picking. The system, which combin…
-
New neural operator framework offers faster, more accurate CT scans
Researchers have developed a new framework called Computed Tomography neural Operator (CTO) that utilizes neural operators to reconstruct images from sparse X-ray projections. Unlike previous methods that overfit to spe…
-
New Aftab architecture boosts deep reinforcement learning efficiency
Researchers have developed "Aftab," a novel architecture for parallelized Q-networks that enhances sample efficiency and representational capacity in deep reinforcement learning. This new framework systematically evalua…
-
MAUPITI enables on-device learning for low-power infrared sensors
Researchers have developed MAUPITI, a novel on-device learning system for low-resolution infrared sensors. This system utilizes a smart multi-pixel IR sensor with a RISC-V microcontroller, enabling pose and gesture reco…
-
New framework analyzes similarity development in vision networks
Researchers have introduced the Deep Similarity Inspector (DSI), a novel framework designed to systematically analyze how similarity perception develops within supervised vision networks during training. This tool allow…
-
Pruning, adversarial training, and hardware faults interact to affect DNN reliability
A new research paper investigates the combined impact of model pruning, adversarial training, and hardware faults on the reliability of deep neural networks. The study found that while adversarial training enhances robu…
-
New 'Neural Echo' Framework Bridges Signal Processing and Explainable AI
Researchers have introduced a new framework called the "neural echo" to better understand the internal workings of neural networks. This method generalizes concepts from classical signal processing, such as impulse resp…
-
AI enhances rip current detection using UAVs and wavelet texture analysis
Researchers have developed a new method for monitoring rip currents using unmanned aerial vehicles (UAVs) by integrating wavelet-derived texture features with deep learning. This approach enhances the detection of subtl…
-
HiResNets enable native Full-HD video recognition with human-like foveation
Researchers have developed HiResNets, a novel approach to video recognition that significantly reduces the computational cost associated with high-resolution inputs. By employing a foveal residual stream and log-polar i…
-
New digital OTA framework boosts federated edge learning for IoT
Researchers have developed a novel digital over-the-air (OTA) computing framework designed to enhance federated edge learning (FEEL) in Internet of Things (IoT) deployments. This new approach jointly trains a random acc…
-
Specialized design data training challenges reliance on massive pretraining
A new research paper explores the effectiveness of pretraining strategies for machine learning models when applied to specialized design data. The study, using the JONES-19 dataset derived from "The Grammar of Ornament,…
-
New framework bridges AI and power engineering education · 2 sources tracked
A new framework, Engineering-Grounded AI (EGAI), has been developed to integrate artificial intelligence into power and energy systems education. This framework, presented as a collection of open, executable Jupyter not…
-
New CENDRe method extracts concepts from CNN time-series models
Researchers have developed CENDRe, a novel concept extraction method designed for convolutional neural networks (CNNs) used in time-series classification. This method addresses limitations of existing techniques by anal…
-
New BCNet framework improves CT-based bronchus classification
Researchers have developed BCNet, a novel framework for classifying bronchi using CT scans. This structure-guided approach integrates segment-level topological information from point clouds with voxel-level representati…
-
Machine learning pilot study shows promise for COVID-19 classification from X-rays
Researchers have conducted a pilot study using traditional machine learning techniques to classify COVID-19 from other pneumonias using chest X-ray data. By employing texture and gradient-based features with classifiers…
-
Deep learning framework accurately classifies plant nitrogen stress · 1 source tracked
Researchers have developed a novel deep learning framework to classify nitrogen stress severity in plants, particularly when combined with other environmental stressors like drought and weed competition. The model integ…
-
Domain Adaptation Techniques Evaluated for Acoustic Scene Classification
This paper investigates domain adaptation techniques for acoustic scene classification, focusing on convolutional neural network (CNN) and transformer-based feature representations. The study evaluates two methods, Doma…
-
CNNs rely on pixel intensity over texture for vascular imaging
A new research paper explores how Convolutional Neural Networks (CNNs) interpret visual information for vascular segmentation in microscopy and fundus imaging. The study found that pixel intensity is more crucial than t…