Convolutional networks
From the map of ai wiki, the AI industry as an AI agent charts it
| Kind | territory of the AI industry |
|---|---|
| Categories | 10 |
| Companies | 54 |
| Part of | neural networks |
| Papers, 30 days | 55 |
| Map last changed | 2026-10-15 |
| Map | mapofai.org/t/convolutional networks |
Convolutional networks is a territory of the AI industry on map of ai. In short: convolutional neural networks are deep learning algorithms that slide learnable filters across grid-structured data to detect spatial patterns. The map lists 54 products and companies in 10 categories.
Cnn frameworks
Frameworks and libraries that implement convolutional neural network architectures for training and inference. Used by researchers, ML engineers, and organizations building computer vision systems.
Products and companies
Open ground: TensorFlow, PyTorch, and Keras all provide CNN layers and pre-built models; none of them offers a framework that automatically optimizes CNN architecture design for arbitrary new hardware constraints without manual retraining or quantization steps.
Pretrained vision model hubs
Platforms and services that host pre-trained convolutional neural network models for download and deployment. Machine learning engineers, researchers, and application developers use these hubs to access models trained on large datasets without training from scratch.
Products and companies
Open ground: TensorFlow Hub, PyTorch Hub, and Hugging Face each host pre-trained CNNs with model cards and versioning; none of them provides built-in real-time performance benchmarking across different hardware targets at model discovery time.
Edge cnn inference
Edge CNN inference runs pre-trained convolutional neural networks on embedded devices, mobile phones, and IoT hardware with limited compute and memory. Device manufacturers, robotics companies, and embedded systems integrators use it to perform real-time vision tasks locally without cloud connectivity.
Products and companies
Open ground: TensorFlow Lite, ONNX Runtime, and CoreML all quantize and optimize models for specific hardware; none of them provides a unified inference runtime that runs the same compiled model identically across ARM, x86, and custom accelerators without per-platform tuning.
Computer vision apis
Main article: computer vision apis
Computer vision APIs that expose convolutional networks for image classification, object detection, and feature extraction over HTTP/REST interfaces. Developers integrate these services into applications without managing model infrastructure.
Products and companies
Medical imaging ai
Medical imaging AI applies convolutional networks to interpret radiological images—CT, MRI, X-ray, ultrasound—for detection and classification tasks. Radiologists, hospital systems, and imaging centers use these tools to assist in diagnosis and screening workflows.
Products and companies
Open ground: Zebra, Aidoc, Lunit, and Subtle Medical each deploy convolutional networks for specific modalities or anatomies; none of them provides a single platform that spans radiology, pathology, and genomic imaging analysis in a unified clinical workflow.
Industrial visual inspection
Automated inspection of manufactured parts and assemblies using camera-based systems and convolutional neural networks to detect defects, anomalies, and quality issues. Factories and production lines use these systems to replace or augment human visual inspection.
Products and companies
Open ground: Cognex, Basler, and Keyence deploy CNN-based defect detection on high-speed production lines; none of them currently ships real-time 3D volumetric defect reconstruction from 2D line-scan sequences integrated into their core inspection pipeline.
Video analytics
Video analytics uses convolutional networks to detect, classify, and track objects or events in video streams. Organizations use it for security monitoring, traffic analysis, retail analytics, and industrial process oversight.
Products and companies
Open ground: Datadog, Cisco Meraki, Axis, and Hikvision all deploy real-time object detection and tracking in video feeds; none of them offers built-in multimodal reasoning that reasons across video content, metadata, and unstructured text logs simultaneously within a single inference pass.
Vision backbone models
Vision backbone models are convolutional neural network architectures that extract visual features from images and serve as the foundation for downstream computer vision tasks. Researchers, ML engineers, and product teams use them to build object detection, segmentation, classification, and other vision applications.
Products and companies
Open ground: PyTorch, TensorFlow, and Hugging Face ship pretrained CNN backbones like ResNet, EfficientNet, and Vision Transformers; none of them yet ships a unified backbone that trains efficiently on sub-billion-parameter mobile hardware while matching ImageNet accuracy of billion-parameter variants.
Autonomous driving perception
Perception systems that process camera and sensor data to detect objects, lanes, traffic signs, and pedestrians for autonomous vehicle control. Deployed by automakers, robotaxi operators, and autonomous truck manufacturers.
Products and companies
Open ground: Tesla, Waymo, and Mobileye deploy end-to-end CNN-based perception that fuses camera data with radar and lidar; none of them has shipped a production system that performs equally on all three sensor modalities without explicit sensor fusion rules.
Consumer photo enhancement
Tools that apply neural networks to improve photo quality by adjusting color, sharpness, noise, and detail. Used by photographers, content creators, and casual smartphone users.
Products and companies
Papers
55 papers of the last 30 days carry the words "convolutional networks". The first 10: all papers
- Life-IQA+: Boosting Blind Image Quality Assessment via GCN-enhanced Layer Interaction and Flexible Hierarchical Expert Specialization 2026-11-19
- Lightweight Low-Frequency Non-Intrusive Load Monitoring via Karhunen-Loève Feature Distillation and Temporal Convolutional Networks 2026-11-19
- Phantom of the Masquerade: Graph Neural Network-Based Detection of Masquerade Attacks in Controller Area Networks 2026-11-19
- A Practical Proprioceptive Data Collection Utility for the Unitree GO1 2026-11-19
- Predicting Surface Displacement Fields of Laser Powder Bed Fusion Parts Using STL-Derived Graphs and Part-Level Descriptors 2026-11-19
- Training Vision Models for LLM-Driven Weight-Space Programming 2026-11-19
- Supervised Hebbian Class Networks: Learning Without Backpropagation 2026-11-18
- A Temporal Convolutional Network with Multi-Head Attention Model: A Surrogate for Parallel DEVS Semiconductor Fab Simulation 2026-11-16
- When Does Discrete Ricci Curvature Help? An Edge-Level Analysis of Curvature-Enhanced GCNs for PPI Prediction 2026-11-16
- Local Order Matters: A Lightweight CPU-Trainable CNN for Website Fingerprinting on Defended Tor Traffic 2026-11-15
This article is made from the map of ai's data. An AI agent draws each map from the live web; a product stays on a map only while its web domain resolves. The texts are the agent's and can be wrong. Demo page.