YOLO12-l is the large variant of the 12th-generation YOLO series, tailored for high-precision visual detection tasks in scenarios with sufficient computing resources. Compared to YOLO12-m, YOLO12-l features a deeper network architecture and wider feature channels, offering stronger representational capacity and contextual understanding—ideal for multi-class detection, occlusion handling, and fine-grained recognition in complex environments.

The model integrates advanced attention mechanisms, an improved cross-scale feature fusion network (FPN + PAN), and optimized loss functions, achieving superior mean Average Precision (mAP) on standard datasets such as COCO and VOC. Despite its larger size, YOLO12-l maintains competitive inference speed, making it suitable for server-side deployment, smart transportation systems, industrial automation, and medical image analysis.

YOLO12-l supports ONNX, TensorRT, OpenVINO, and other mainstream inference frameworks, and is deployable across multiple platforms. It is a robust choice for production-level applications that demand high accuracy and reliability.

Source model

• Input shape: 1x3x640x640
• Number of parameters: 25.22M
• Model size: 101.11M
• Output shape: 1x84x8400

Source model repository: yolo12l

When the user has fine-tuned the source model, the model conversion process must be performed again.

Users can refer to either of the following two methods to complete the model conversion:

• Using AIMO for model conversion: Click Model Conversion Reference in the Performance Reference section on the right to view the conversion steps.

• Using Qualcomm QNN for model conversion: Please refer to the Qualcomm QNN Documentation.

The model performance benchmarks and example code provided by Model Farm are all implemented based on the APLUX AidLite SDK.

For models in .bin format, you can use either of the following two inference engines to run inference on Qualcomm chips:

Inference using APLUX AidLite: please refer to the APLUX AidLite Developer Documentation

Inference using Qualcomm QNN: Please refer to the Qualcomm QNN Documentation

Inference Example Code

The inference example code is implemented using the AidLite SDK.

Click Model & Code to download the model files and the inference code package. The file structure is as follows:

/model_farm_{model_name}_aidlite
    
    |__ models # folder where model files are stored

    |__ python # aidlite python model inference example

    |__ cpp # aidlite cpp model inference example

    |__ README.md