Images generated by AI makes it hard to differentiate what’s real and fake

9

October

2018

No ratings yet.

People are now aware of the technological changes due to artificial Intelligent (AI). So far, human can easily tell the difference between real images and fake images created by AI since these fake images are pretty strange to human eyes. However, with the advancement of AI technology, it is getting harder and harder to spot the slight difference between these two different sources of images.

Recently, the team formed by researchers from DeepMind and Heriot-Watt University in the UK created a machine learning model, BigGANs, which are based on generative adversarial network (GAN). BigGANs are stated to improve the quality of images generated by the AI image generator.

During the training process, BigGANs rely on ImageNet, an image dataset that contains numerous images of different objects and is maintained by Stanford and Princeton. Trained on 128 Google’s Tensor Processing Units (TPUs), the model required one to two days to finish training. The results measured by inception score (IS) showed that the model worked well by pushing IS from 52.52 to 166.3. It is believed that the reason why BigGANs become a success is because they employ larger GAN, use bigger batch sizes and involve more parameters.

One might be curious about why this application of AI is important. Fake images might cause a lot of problems, such as privacy, morality, legal issues, etc.; however, image generators are very useful for model training, especially when only limited training data is available. Under normal circumstances, models perform better with more diversified training data. Therefore, if theses image generators can provide different but realistic images, they can alleviate problems of the lack of training data. Moreover, to avoid misappropriated uses of GANs technology which can be led by political or unethical purposes, the research team focuses on more general images instead of images with faces.

 

Sources:

https://venturebeat.com/2018/10/02/deepmind-ai-can-generate-convincing-photos-of-burgers-dogs-and-butterflies/

https://www.theregister.co.uk/2018/10/01/biggan_fake_images/

Please rate this

Just talk and buy: the current state and the future of voice shopping

11

September

2018

No ratings yet.

Online shopping has become a popular trend. With a smart device, you can shop without leaving your house. It could be even more convenient when you shout out what you need and within a few days, your order arrives at your place. In this case, it doesn’t even require your hands to press any buttons. This way of shopping is termed as voice shopping, which indicates purchases made with smart speakers, and is possible now due to advances in speech recognition. With the rise of artificial intelligence and big data, the accuracy rate in speech recognition technique has improved and the technique is considered mature enough to permit its use as designed.
Currently, Amazon’s Alexa Voice Shopping provides a service that enables customers to make purchases and check orders with just voices through smart speakers, such as Amazon Echo. A recent market research by OC&C Strategy Consultants predicts a massive increase in voice shopping. The research estimates that the industry will grow from $2 billion today to $40 billion in 2022 driven by the growing popularity of smart speakers. However, according to the report of The Information, a digital media company, only two percent of Amazon Alexa owners have experienced voice shopping and among these people, more than 90% have only tried it once.
The fact that people are not used to voice shopping is obvious and is an example that the technology runs ahead of the business model. To be more specific, the speech recognition technology has changed the way to shop, but the changes are beyond people’s imagination. Therefore, besides providing products and services, companies have to enhance customers’ motivation in changing shopping ways and let customers feel the benefits of voice shopping to gain success in voice shopping industry, which are not an easy task. There’s still a long way to go before reaching $40 billion estimate.

References:
Sorry, Alexa – practically no one uses you to buy things from Amazon
https://www.theinformation.com/articles/the-reality-behind-voice-shopping-hype
https://www.tomsguide.com/us/alexa-voice-shopping-tutorial,news-25370.html

Please rate this