#MobileNetV2
#MAV
Fusing STFT and Mel spectrograms through a CBAM-enhanced MobileNetV2 lets a Tello quadrotor detect propeller damage from in-flight audio at low compute cost.
Details: https://www.maxapress.com/article/doi/10.48130/mav-0026-0015
#FaultDiagnosis #MobileNetV2 #AudioSignal
November 6, 2026 at 4:15 AM
Seeking Independent Reviewers: Wheat Rust AI Cross-Dataset Generalisation
Hello Hugging Face community! I’m Devanandha VS, an independent AI researcher and MSc Artificial Intelligence graduate from Ulster University, UK. I’m seeking independent researchers, computer-vision practitioners and open-source developers interested in evaluating an agricultural AI research project. ## Project Overview **Wheat Rust AI Detection** is an open-source deep-learning project using a MobileNetV2-based classifier to identify three wheat leaf categories: * Healthy * Brown Rust * Yellow Rust The classifier achieved **98.78% accuracy (728/737 correct predictions)** on its internal validation dataset. However, subsequent evaluations using independent datasets revealed substantial generalisation weaknesses. Rather than presenting the initial accuracy as proof of real-world reliability, I investigated and publicly documented the model’s limitations, unsuccessful experiments and Grad-CAM explainability findings. The project is an experimental research demonstration, not a production-ready agricultural diagnostic system. ## What I’m Looking For I welcome independent contributions in four areas: **1. Cross-Dataset Generalisation** Evaluate the model on independent wheat leaf datasets and investigate differences in performance across data sources. **2. Reproducibility** Attempt to reproduce the documented experiments, including preprocessing, evaluation methodology and dataset configurations. **3. Explainable AI** Review the Grad-CAM analysis and assess the usefulness and limitations of the generated explanations. **4. Model Reliability** Explore improved evaluation methods, relevant baselines, calibration techniques and strategies for addressing dataset shift. ## How to Participate I’ve opened a dedicated GitHub issue for independent evaluation and collaboration: **Independent Reproduction and External Validation:** Open issue devanandha/wheat-rust-ai-detection #### Call for independent reproduction and external validation — Wheat Rust AI opened 12:44PM - 05 Nov 26 UTC devanandha I'm inviting independent researchers, machine-learning practitioners, plant-path…ology specialists, and open-source developers to **evaluate the Wheat Rust AI Detection project critically**. ### Background The project's MobileNetV2 classifier achieved **98.78% accuracy (728/737)** on its **internal validation set**, but later experiments using independent datasets revealed **substantial generalisation limitations**. This project is a research demonstration, **not a production-ready agricultural diagnostic tool**. I welcome positive and negative findings alike. ### Areas where external contributions would be useful 1. **Reproducibility:** Attempt to reproduce documented evaluation results; report environment, software versions, dataset provenance, splits, preprocessing and any discrepancies. 2. **External validation:** Evaluate on suitably licensed independent wheat-leaf image sets, with transparent class mapping and attention to data leakage. 3. **Dataset shift:** Examine possible reasons for performance differences across sources, such as acquisition conditions, label definitions or class imbalance. 4. **Explainability:** Critically assess Grad-CAM visualisations and interpretation, including potential limitations. 5. **Improved methods:** Propose or test baselines, uncertainty estimates, calibration methods and robustness evaluations. ### Where to contribute - **GitHub repository:** https://github.com/devanandha/wheat-rust-ai-detection - **Archived project:** https://doi.org/10.5281/zenodo.22726591 - **Interactive research demonstration:** https://devanandha-wheat-rust-ai.streamlit.app/ Please share independent observations in a comment, open a dedicated issue for a specific result, or submit a pull request for reproducible improvements. Include sufficient methodological detail for others to review your findings. Respect third-party dataset licences and privacy; avoid uploading data that cannot be redistributed. **There is no expectation of a favourable review, endorsement or citation.** Critical feedback, unsuccessful reproduction attempts and alternative explanations are equally welcome. The goal is to improve transparency, reproducibility and reliability in agricultural computer vision. **Project Repository:** Repository devanandha/wheat-rust-ai-detection ### GitHub - devanandha/wheat-rust-ai-detection: AI-powered wheat rust classification using... AI-powered wheat rust classification using MobileNetV2, with reproducible evaluation and a Streamlit demonstration. Contributors are welcome to report findings, suggest improvements, submit pull requests or discuss potential methodological limitations. Positive feedback, critical assessments and unsuccessful reproduction attempts are equally welcome. ## A Question for the Community When evaluating agricultural image classifiers across independent datasets, what factors would you investigate first? * Differences in image acquisition and environmental conditions? * Class definitions and annotation quality? * Data preprocessing and dataset composition? * Domain shift, class imbalance or model calibration? I’d appreciate hearing your experience and welcome collaboration on a more rigorous evaluation. Thank you for your time and consideration! **Devanandha VS** MSc Artificial Intelligence, Ulster University Independent AI Researcher
discuss.huggingface.co
November 5, 2026 at 2:33 PM
MobileNetV2's 3.5M params and ~72% top-1 remain a solid edge baseline — but this card has no weights format, no licence beyond "other", and an unverified `amd` uploader. Hard to see it beating a documented ONNX export for production until provenance is clear.
Tiny but mighty: MobileNetV2 lands on Modelers.cn for Ascend edge vision
Modelers.cn, the Huawei/Ascend AI ecosystem platform, has added a repository for MobileNetV2 under the account amd. If you build computer vision for phones, cameras, robots or other constrained…
aichina.news
October 24, 2026 at 9:06 PM
Quantize MobileNetV2 to INT8 naively and ImageNet accuracy falls from 71.7% to 0.1%. Data-free repairs bring it back to 71.2%.

Yet one unsupported op can leave the engine slower than FP16.

TensorRT engineers: which ops keep your engines from staying INT8?

#EdgeAI #Jetson
October 24, 2026 at 11:27 AM
ExecuTorch is now an official Zephyr external module. Zephyr Project’s walkthrough covers setup and runs quantized MobileNetV2 on an Alif Ensemble E8 DevKit.

https://www.zephyrproject.org/executorch-integrated-with-zephyr-rtos/
ExecuTorch Integrated with Zephyr RTOS
ExecuTorch is now a Zephyr module. Add it to your manifest, configure it in prj.conf, build with west, and run a quantized MobileNetV2 on an Arm Ethos-U55 NPU.
www.zephyrproject.org
October 17, 2026 at 12:30 AM
A network can lose 90% of its weights and run at the same speed: dense kernels still multiply every zero.

Cut 70% of a MobileNetV2 block's channels and 70% of its MACs go too.

Jetson engineers: do you ship 2:4 sparsity, or still prune channels?

#EdgeAI #Jetson
October 16, 2026 at 12:01 PM
MobileViT-XS has a third fewer parameters than MobileNetV2 and 1.5 points more accuracy. On an iPhone 12 it is about 8x slower.

Parameters said faster. MACs said 2.3x. The phone said 7.9x.

Runtime engineers: has fused attention closed that gap on Jetson?

#EdgeAI #Jetson
October 9, 2026 at 11:47 AM
🚀 FeliniAI — Detector de Alergias Felinas

Sistema de diagnóstico asistido de alergias en gatos mediante triple pipeline de IA: visión computacional con MobileNetV2 (imágenes reales de datasets públicos), clasificador c…

👉 https://adrianmoreno-dev.com/proyecto/feliniai

#IA #MachineLearning #Salud
October 3, 2026 at 11:00 AM
Automated Detection of Soil-Transmitted Helminths and Schistosomiasis with Mobile Deployment of a Quantized MobileNetV2 Model from microscopic images #NeuroDegeneration 🧪🧠
https://www.researchsquare.com/article/rs-11016494/latest
September 24, 2026 at 4:02 PM
MobileNetV2 has 47% fewer MACs than MobileNetV1. On the @nvidiabot.bsky.social Jetson Orin Nano it runs about 7% faster.

Every efficiency trick relocates the bottleneck: to the 1x1 then to memory.

Kernel and compiler engineers: has fusion closed the depthwise gap on your hardware?

#EdgeAI #Jetson
September 24, 2026 at 1:30 PM
Transfer Learning with MobileNetV2 in Keras: A Practical Guide
How I Built a Flower Classifier by Borrowing a Pretrained Brain A practical guide to transfer learning …

https://pub.towardsai.net/transfer-learning-with-mobilenetv2-in-keras-a-practical-guide-f5cc4935108a?source=rss----98111c9905da---4
September 22, 2026 at 7:29 PM
MobileNetV2 on the Jetson Orin Nano: 12.5 ms end to end. Convolution is 3.8 ms of it.

Time only the kernels and you are 3x optimistic. Skip the synchronize and you are 100x optimistic.

What fraction of your reported latency is actually the model?

#EdgeAI #Jetson
September 5, 2026 at 7:01 PM
Обзор статьи MobileNetV2: Умный крошечный гигант

Понимание и реализация MobileNetV2 с PyTorch — следующее поколение MobileNetV1

#ai #aipaper #news
MobileNetV2 Paper Walkthrough: The Smarter Tiny Giant
towardsdatascience.com
October 4, 2025 at 1:19 PM
Изучение эффективности классификации изображений с помощью MobileNetV2

MobileNet - это модель с открытым исходным кодом, созданная для поддержки разработки смартфонов. В ней используется архитектура сверточной нейронной сети (CNN) для выполнения задач компьютерного зрения, таких как кла…

#ai #news
Exploring the Efficiency of Image Classification With MobileNetV2
www.analyticsvidhya.com
February 26, 2025 at 12:17 PM
Learn how to implement the MobileNetV2 architecture from scratch in PyTorch. In this new article, Muhammad Ardi explains the key concepts from the original paper, including inverted residuals and linear bottlenecks, and provides a complete, step-by-step coding guide.
MobileNetV2 Paper Walkthrough: The Smarter Tiny Giant | Towards Data Science
Understanding and implementing MobileNetV2 with PyTorch  — the next generation of MobileNetV1
towardsdatascience.com
October 11, 2025 at 3:27 PM
Learn how to implement the MobileNetV2 architecture from scratch in @PyTorch.org. Muhammad Ardi explains the key concepts from the original paper, including inverted residuals and linear bottlenecks, and provides a complete, step-by-step coding guide.
MobileNetV2 Paper Walkthrough: The Smarter Tiny Giant | Towards Data Science
Understanding and implementing MobileNetV2 with PyTorch  — the next generation of MobileNetV1
towardsdatascience.com
October 9, 2025 at 6:18 PM
Learn how to implement the MobileNetV2 architecture from scratch in PyTorch. In this new article, Muhammad Ardi explains the key concepts from the original paper, including inverted residuals and linear bottlenecks, and provides a complete, step-by-step coding guide.
MobileNetV2 Paper Walkthrough: The Smarter Tiny Giant | Towards Data Science
Understanding and implementing MobileNetV2 with PyTorch  — the next generation of MobileNetV1
towardsdatascience.com
October 3, 2025 at 9:05 PM
MobileNetV2, on the ImageNet dataset, demonstrating that the generated architectures can deliver up to 440 GOPS/W with relatively small output error during inference, outperforming several State-of-the-Art CGRA architectures in terms of throughput and [7/8 of https://arxiv.org/abs/2505.23553v1]
May 30, 2025 at 5:54 AM
dynamic for different iterations to traverse most of the parameters in the update layers along the time dimension for better performance. The experimental result shows that the proposed method enables an ImageNet pre-trained MobileNetV2 trained on [4/5 of https://arxiv.org/abs/2503.17959v1]
March 25, 2025 at 5:54 AM