A Talent Dev of deep learning: Mr. Wang Ha

Published on 3 August 2023 at 05:35

With over 15 years of experience, I have delivered numerous projects and studies that showcase my skills and expertise:

 

➡️ Developed a Generative Adversarial Network (GAN) based facial image augmentation approach as part of my PhD Thesis. The approach used latent space exploration via GAN inversion for acquiring median faces, and achieved better than GAN based approaches in the literature.

 

➡️Performed facial expression recognition of 3D facial expression datasets, utilizing both point clouds and surface patches to improve accuracy on emotion recognition. The approach achieved better recognition accuracy by combining these methods.

 

➡️Developed an eye state detection framework for driver fatigue detection as a part of my master's thesis. The approach used a Haar cascade approach for face and eye localization, while circular Hough detection algorithm was used to detect eye openness state. This approach is able to alert on driver fatigue.

 

➡️Developed an LSTM-based stock price forecaster using features derived from candlesticks. The features were formed with the help of an expert mathematician using 5m, 10m, 1h, 4h, and 1d candle data.

➡️I have also worked on projects in Natural Language Processing. My technical expertise includes proficiency in the following languages: ➡Python, C, C#, MATLAB I am skilled in DL frameworks such as,

 

➡️Tensorflow, Pytorch, Keras and utilize libraries like

 

➡️OpenCV, NumPy, SciPy, Scikit-learn, Pandas, and Matplotlib to develop my solutions.

If you need an experienced and knowledgeable computer vision expert to help you with your project, feel free to reach out to me. I am always ready to take on new challenges and deliver quality results.

 

 

 

凭借超过 15 年的经验,我完成了许多项目和研究,展示了我的技能和专业知识:

 

➡️ 开发了一种基于生成对抗网络(GAN)的面部图像增强方法,作为我博士论文的一部分。 该方法通过 GAN 反演使用潜在空间探索来获取中值面,并且比文献中基于 GAN 的方法取得了更好的效果。

 

➡️对3D面部表情数据集进行面部表情识别,利用点云和表面补丁来提高情绪识别的准确性。 该方法通过结合这些方法获得了更好的识别精度。

 

➡️开发了一个用于驾驶员疲劳检测的眼睛状态检测框架,作为我硕士论文的一部分。 该方法使用 Haar 级联方法进行面部和眼睛定位,同时使用圆形 Hough 检测算法来检测眼睛张开状态。 这种方法能够对驾驶员疲劳发出警报。

 

➡️使用源自烛台的特征开发了基于 LSTM 的股票价格预测器。 这些特征是在专家数学家的帮助下使用 5m、10m、1h、4h 和 1d 蜡烛数据形成的。

 

➡️我还参与过自然语言处理的项目。 我的技术专长包括精通以下语言:➡Python、C、C#、MATLAB 我擅长 DL 框架,例如:

➡️Tensorflow、Pytorch、Keras 并利用诸如

 

➡️OpenCV、NumPy、SciPy、Scikit-learn、Pandas 和 Matplotlib 来开发我的解决方案。

 

如果您需要经验丰富且知识渊博的计算机视觉专家来帮助您完成项目,请随时与我联系。 我随时准备接受新的挑战并交付高质量的成果。

Published on 30 January 2023 at 05:35

 

Add comment

Comments

There are no comments yet.