Shuo Cheng

Title: Innovations of Shuo Cheng

Introduction

Shuo Cheng is a prominent inventor based in Los Angeles, CA. He has made significant contributions to the field of neural networks and object detection, holding a total of four patents. His work focuses on enhancing the capabilities of artificial intelligence through innovative systems and methods.

Latest Patents

One of Shuo Cheng's latest patents is titled "Systems for multi-task joint training of neural networks using multi-label datasets." This invention provides systems and methods for training a neural network that includes an encoder module and a multi-headed attention mechanism. The system is designed to receive input data with multiple sets of labels, extract features, and compute training loss metrics for each set of labels. This innovative approach allows for more efficient training of neural networks.

Another notable patent is "Box detection for object attachment." This patent describes techniques for determining a bounding box around objects in images. The process involves estimating frames from the image using a neural network, detecting planar regions, and performing vanishing point detection. The results are fused to determine the depth value of each pixel, ultimately allowing for accurate bounding box location determination.

Career Highlights

Shuo Cheng is currently employed at Lemon Inc., where he continues to develop cutting-edge technologies in the field of artificial intelligence. His work has garnered attention for its practical applications and innovative methodologies.

Collaborations

Shuo Cheng collaborates with talented individuals such as Wanchun Ma and Linjie Luo. Together, they contribute to advancing the field of neural networks and object detection.

Conclusion

Shuo Cheng's innovative patents and contributions to artificial intelligence demonstrate his expertise and commitment to advancing technology. His work continues to influence the development of intelligent systems that enhance various applications.