Run F5-TTS voice cloning from .NET with pure ONNX Runtime — no Python. FP16, any-length text, German & English models.

OCR + ONNX Runtime

Currently supporting VOXCPM 1.5B

Standalone C++ inference project for VoxCPM models built on top of ggml.

用 onnx 和 gguf 格式混合运行 Fun-ASR-Nano 模型全流程

A collection of ONNX models for Dolphin Automatic Speech Recognition by DataoceanAI and Tsinghua University.

Collect common processes in onnx format.

Convert the SOTA driable area and lane detection models into ONNX format.

Convert the SOTA depth models into ONNX format.

A novel Multimodal Large Language Model (MLLM) architecture, designed to structurally align visual and textual embeddings.

Utilizes ONNX Runtime for TTS model.

Utilizes ONNX Runtime to get the sentence embedding vectors and similarity.

Transcribe subtitles and translate them offline with ease.

A concise guide on using ONNX Runtime Execution Providers to optimize model performance across various hardware platforms.

Export the STFT or ISTFT process in ONNX format.

Uses ONNX Runtime for character role speaker identification.

Utilizes ONNX Runtime to transcribe audio into text.

Utilizes ONNX Runtime for speech activity detection.

Utilizes ONNX Runtime for audio denoising.

Running the F5-TTS by ONNX Runtime standalone with GUI

A Toolkit to Help Optimize Onnx Model

Demonstrate Yolov9 model with Qualcomm Hexagon NPU and DirectML

Running the F5-TTS by ONNX Runtime

Demonstration of combine YOLO and depth estimation on Android device.

Demonstration of running a native LLM on Android device.