This page list the tools contained by the AI Toolbox.
| Tool name | Short description | Integration |
|---|---|---|
| LLM Tool | Large Language Models as deployable services from the HuggingFace model hub | deep |
| Object detection tool | Object detection tool based on Ultralytics models | deep |
| Demo Tool | Example tool for presenting deep integration int the AI Toolbox | deep |
| Prompt Context Injection Tool | Injecting context into NLP prompts | deep |
| 6DoF Pose Estimation Tool | 6 Degrees-of-Freedom pose estimation for robotic applications | shallow |
| Object detection with YoloV8 | Real-time object detection using YOLOv8 architecture | shallow |
| Image segmentation detection with BiRefNet | High-accuracy image segmentation using BiRefNet backbone | shallow |
| Optical Character Recognition with TrOCR | Text recognition from images using Transformer-based OCR | shallow |
| Large Language Model using DeepSeek-R1-0528-Qwen3-8B | Deployment-ready DeepSeek-R1-0528-Qwen3-8B LLM for advanced language tasks | shallow |
| Zero Shot Object Detection with OWLv2 | Object detection without prior training using OWLv2 models | shallow |
| Depth-based Isolation Forest Feature Importance | Interpretable Anomaly Detection with DIFFI | shallow |
| Extended Isolation Forest Feature Importance | Interpretable Anomaly Detection with ExIFFI | shallow |
| Mobile Segment Anything | Mobile Segment Anything | shallow |
| Segment Anything Model 3 | SAM 3 is an AI model that segments and tracks every instance of a concept across images and videos using text or image prompts. | shallow |
| Real-Time Detection Transformer (RT-DETR) | RT-DETR is a cutting-edge end-to-end object detector that provides real-time performance while maintaining high accuracy | shallow |
| Foundation Pose | Unified 6D Pose Estimation and Tracking of Novel Objects | shallow |
| DeepSeek-OCR | Contexts Optical Compression | shallow |
| Qwen3.5 | Qwen3.5 is a high-efficiency multimodal foundation model featuring hybrid architecture and advanced reasoning across 201 languages. | shallow |
| GPT-OSS 120B | OpenAI’s open-weight models designed for powerful reasoning, agentic tasks, and versatile developer use cases. | shallow |
| Tabular Foundation Models | Foundation Model for Tabular Data | shallow |
| Lag-Llama | Open-source foundation model for time series forecasting | shallow |
| LayoutLMv3 | A unified Document AI model using ViT-style patches and joint masking to achieve state-of-the-art document understanding. | shallow |
| Embedding Tool | Text embedding generation for semantic search, clustering, retrieval, and similarity tasks | shallow |
| Reranker Tool | Document reranking for improving retrieval precision in search and RAG pipelines | shallow |
| Speech-to-Text Tool | Automatic speech recognition and speech translation for multilingual audio transcription | shallow |
| Text-to-Speech Tool | Multilingual text-to-speech and voice cloning for natural speech generation | shallow |
| Document Parsing Tool | Structured document conversion and PDF understanding with layout, table, and content extraction | shallow |
| Image Classification Tool | Image classification for assigning category labels to images without object localization | shallow |
| Multi-Object Tracking Tool | Real-time multi-object tracking across video frames for surveillance, robotics, and analytics | shallow |
| Human Pose Estimation Tool | Human keypoint and skeletal pose estimation for movement analysis and interaction understanding | shallow |
| Multimodal VLM Tool | Vision-language reasoning over images and text for captioning, visual question answering, and multimodal assistance | shallow |
| Time-Series Anomaly Detection Tool | Detection of unusual temporal patterns and anomalies in sensor, finance, and operational time series | shallow |
| Tabular Prediction Tool | Classification and regression on structured tabular datasets for practical machine learning applications | shallow |
| Explainability Tool | Model interpretability and explanation for tabular, vision, and deep learning systems | shallow |
| Device-Level Quality Detection Tool | Detection of device-level defects and linkage of machine-signal-related features to root causes | shallow |
| FFT feature extraction for tool breakage detection | A comprehensive methodology for detecting tool breakage in milling operations through frequency domain analysis of vibration signals. | shallow |
| Factual Correctness Evaluation of AI-Generated Summaries | This tool evaluates the factual correctness of AI-generated summaries using structured fact representations. | shallow |
| Automatic Ontology Learning Framework | This tool supports ontology learning from a domain corpus or structured database using a lightweight encoder-based pipeline. | shallow |
| NL2SPARQL Semantic Error Detection Framework | A framework and benchmark to test LLMs' abilities to detect errors in SPARQL queries. | shallow |
| GAP-FL | Group-Aware Fairness Analysis in Differentially Private Federated Learning. | shallow |
| ZebraLogicReasoner | Reproducible Reasoning Evaluation of Structured Logical Puzzles. | shallow |
| Motion Playground | A tool to model and visualize motion data in real time. | shallow |
| Phase Retrieval algorithm package | FourierGSNet, GS direct unrolling, Gradient Descent fitting method, and Newton-Raphson fitting method. | shallow |
| Invertible differentiable simulation | Invertible differentiable batched GPU-accelerated simulation for the laser beam shaping system of EOS M290 PBF-LB/M system. | shallow |
| AMR Travel Time Prediction | A machine learning pipeline for predicting AMR travel times from fleet log data. | shallow |
| Natural-Language Programming of Industrial Robots | An AI-powered programming environment that turns a brief natural-language description of a pick-and-place task into a tested, deployable robot control program. | shallow |
| Differentiable-Physics Motion Planning | An AI training method for industrial robot arms that learns smooth, accurate motion trajectories from a physics simulator with full gradient backpropagation. | shallow |