a case study of audio/image aerial scene classification — Multimodal Scene Recognition with image/audio/video as inputs presents challenges in integrating data, designing models, and identifying co-training strategies, in contrast to developing solutions for Single-Modality (e.g., image-only or audio-only) Scene Recognition. This post consists of two parts. The first part provides five multimodal scene recognition datasets for reference. These datasets…

From SMILES to Word Embedding and Graph Embedding — Machine learning has been applied to many problems in cheminformatics and life science, for example, investigating molecular property and developing new drugs. One critical issue in the problem-solving pipeline for these applications is to select a proper molecular representation that featurizes the target dataset and serves the downstream model. Figure…

Deep learning for object detection on image and video has become more accessible to practitioners and programmers recently. One reason for this trend is the introduction of new software libraries, for example, TensorFlow Object Detection API, OpenCV Deep Neural Network Module, and ImageAI. These libraries have one thing in common…

a brief review on deep learning in RapidMiner and Orange — I ran into two visual programming environments for data science and machine learning recently: RapidMiner and Orange. Both systems allow users to build a data science or machine learning solution in a LEGO-like style, i.e., drag-and-drop components to construct a process including data preparation, modeling, evaluation, validation, visualization, etc. …

Google 有沒被 Microsoft 追著打?! — 前幾天聽了中研院資科所研究員古倫維教授的線上演講 ChatGPT 到底在夯什麼, 演講完畢有位聽眾的問題很有趣, 問題大意是: Google 人工智慧不是很厲害嗎? Google 先做出了 AlphaGo, 怎麼現在會被 Microsoft 追著打? 問題太有趣了, 在此粗略發表一下個人的看法: AlphaGo 是 DeepMind 做的, 而不是 Google 做的, 雖然 DeepMind 算是 Google 旗下一員, 但是 AlphaGo 功勞應該算 DeepMind 而不是 Google。 ChatGPT 是 OpenAI 做的, 而不是 Microsoft 做的, 雖然 Microsoft 有過注資 OpenAI, 但是 ChatGPT 功勞應該算 OpenAI 而不是 Microsoft。 Google 好像沒有什麼項目吸收及商用 DeepMind AlphaGo 技術成果, 而且 Google 和 DeepMind 似乎競爭關係大於合作關係。

Q1: What is the difference between ChatGPT and LaMDA? Q2: Who develops LaMDA? Q3: Does Google creates LaMDA? Q4: Is LaMDA sentient? Q5: Are you sure that OpenAI creates LaMDA? Q6: Who develops LaMDA? Q7: Why do you make mistake that LaMDA is developed by OpenAI?

Q1: Would you kill one person to save five people in the trolley problem? Q2: What kind of training data is related to your answers on the trolley problem? Q3: What if the trolley problem happens in the self-driver car, should the car kill one person to save five people? Q4: List three ethical problems in a self-driving car. Q5: List three major problems if prioritize the safety of the driver over the safety of the passengers or other road users.

Reference

A Case Study of Defect Detection on Printed Circuit Boards — Automated optical inspection is a common process used in the electronics or manufacturing industry to identify product defects. Conceptually, all the practices in deep learning for computer vision, including image classification, object detection, and semantic segmentation, could be used in automated optical inspection. This post describes a case study of…