MobileNetV3
PulseAugur coverage of MobileNetV3 — every cluster mentioning MobileNetV3 across labs, papers, and developer communities, ranked by signal.
4 day(s) with sentiment data
-
GPU Undervolting Boosts CNN Adversarial Robustness and Energy Efficiency
Researchers have developed a method to enhance the adversarial robustness of Convolutional Neural Networks (CNNs) by undervolting their GPUs during training. This technique introduces stochastic perturbations that act a…
-
Flutter developer seeks help with TFLite model errors after image resizing
A user is encountering significant errors when integrating a TensorFlow Lite (TFLite) model, originally trained using MobileNetv3, into a Flutter application. The issue arises after the camera stream captures images, wh…
-
Winning submission details vocal imitation sound querying strategies
Researchers have detailed their winning submission for the AES AIMLA 2025 Challenge, which focused on querying sound effects through vocal imitation. Their approach involved two fine-tuning strategies: contrastive learn…
-
Deep learning models benchmarked for lung cancer histopathology analysis
Researchers have developed a two-stage deep learning framework for analyzing lung cancer histopathology images. The framework systematically compares state-of-the-art architectures for both tissue classification and reg…
-
New AI framework improves catheter and tube placement assessment in X-rays
Researchers have developed UCompCXR, a novel framework designed to improve the accuracy and safety of assessing catheter and tube placement in chest X-rays. This system addresses limitations of current deep learning met…
-
OrthKD framework extracts clinical knowledge for lightweight diabetic retinopathy screening
Researchers have developed OrthKD, a novel knowledge distillation framework designed to extract generalized clinical knowledge from heterogeneous AI models for lightweight deployment in diabetic retinopathy screening. T…
-
Hybrid deep learning model enhances slate tile traceability and classification
Researchers have developed a hybrid deep learning model to improve the traceability and classification of industrial slate tiles. This approach combines feature matching using XFeat and LightGlue with a MobileNetV3-base…
-
Hybrid deep learning model enhances slate tile traceability and classification
Researchers have developed a hybrid deep learning model to improve the traceability and classification of industrial slate tiles. This approach combines instance-aware re-identification and extraction site classificatio…
-
Automated pipeline uncovers bias in MoE4 architecture search
Researchers have developed an automated pipeline to explore heterogeneous 4-Expert Mixture-of-Experts (MoE4) architectures within the LEMUR dataset ecosystem. This pipeline systematically combines base architecture fami…
-
New GazeLNN model predicts human attention for robot navigation
Researchers have developed GazeLNN, a novel and computationally efficient model for predicting human visual attention in real-time. This model utilizes Liquid Neural Networks and MobileNetV3 to predict fixation heatmaps…
-
New Finetuning Method Adapts DNNs for ReRAM In-Memory Computing
Researchers have developed a new finetuning method to adapt deep neural networks for deployment on ReRAM-based in-memory computing hardware. This approach addresses the challenges of I-V non-linearity and retention erro…
-
Stable Diffusion used for data augmentation in indoor scene recognition
Researchers have proposed a new method for indoor scene recognition by using Stable Diffusion to generate synthetic images for data augmentation. This approach addresses the scarcity of training data for indoor environm…
-
Spacecraft Perception Model Achieves Top Ranking in SPARK 2026 Challenge
Researchers have developed a novel segmentation-based detection method for multi-task spacecraft perception, addressing challenges like limited annotated data and difficult visual conditions. Their compact architecture,…
-
LLMs Evaluate AI Explainability in Skin Disease Diagnosis
Researchers have developed a new framework to evaluate the explainability of AI models used for diagnosing facial skin diseases. This framework utilizes large language models (LLMs) like GPT-5.5, Gemini 3.5 Flash, and C…
-
New framework evaluates AI driver models on more than just accuracy
Researchers have introduced a new framework for evaluating driver monitoring models, moving beyond simple accuracy metrics. The Human-Centered Benchmarking Framework (HCBF) assesses models on accuracy, explainability, e…
-
AI distills multiplexed microscopy data for single-channel tissue segmentation
Researchers have developed a cross-modal knowledge distillation framework to improve single-channel tissue segmentation in microscopy. This method transfers knowledge from a foundation model trained on multiplexed image…