Active Hackathon

Baidu Goes On A Patent Frenzy; Applies For ML-Based Audio Synthesis Ownership

Baidu has come out on top as the leading artificial intelligence patent application leader eclipsing the likes of Tencent and Huawei. Reportedly, Baidu is also leading in the highly competitive area of intelligent driving, with 1,237 patent applications. 


Sign up for your weekly dose of what's up in emerging technology.

After years of research, Baidu has developed a comprehensive AI ecosystem and is now at the forefront of the global AI industry.

“Baidu retained the top spot for AI patent applications in China because of our continuous research and investment in developing AI, as well as our strategic focus on patents.” -Victor Liang, VP, Baidu

via Baidu

Baidu’s patents cover a wide variety of domains that include 

  • Deep learning (1,429 patents)
  • NLP (938 patents) 
  • Speech recognition (933 patents)

While Baidu topped the charts in China, its R&D centre in the US also had applied for patents in the US patent office.

Especially in the speech recognition domain, Baidu has its eyes locked on audio synthesis using CNN. 

In this patent US20190355347A1 for a computer-implemented method for training a neural network model for spectrogram inversion comprising titled ‘Spectrogram to waveform synthesis using convolutional networks’, it lists the following:

  1. Inputting an input spectrogram comprising a number of frequency channels into a convolution neural networks (CNNs)
  2. Outputting from the CNN a synthesised waveform for the input spectrogram, the input spectrogram having a corresponding ground truth waveform
  3. Using the corresponding ground truth waveform, the synthesised waveform, and a loss function comprising at least one or more loss components selected from spectral convergence loss
  4. Using the loss to update the CNN

There is a clear mention of using the convolutional neural networks and since CNNs is the lifeblood of many modern-day ML applications, any claim even on a minor part can hurt in the long run. 

Perils Of A Patent Race

The year 2019 witnessed a sudden growth of interest to own algorithms. So far, Google got a bad rap for going after batch normalisation, a widely used technique in deep learning. Even if the intentions are to safeguard the research from falling to pseudo players, this whole ordeal is a slippery slope where the owners can leverage the smaller firms using advanced technology.

In the case of Baidu, too, there is a danger of losing ownership to many audio processing applications. Baidu is a Chinese company who has contributed to the growing fears amongst the ML community. Its AI vision was fortified with projects like Apollo, an open source autonomous driving platform along with other intelligent driving innovations.

China has allegedly been involved in many intellectual property thefts from US companies. So, when Baidu’s foreign division, applies for a patent, one cannot help but think about the consequences of handing over the ownership to China continued to be the world’s leading source of counterfeit goods, reflecting its failure to take decisive action to curb the widespread manufacture, domestic scale, and export of counterfeit goods.

According to a 2019 United States Trade report, China continues to be the world’s leading source of counterfeit goods responsible for widespread manufacture and export of counterfeit goods.

The most important thing any country can do in the current era is protect its trade secrets. Even though Google has also been accused of patent race, adding China to the Baidu equations changes everything.

The Chinese have a system that encourages transaction of intellectual property allowing every common guy to access cutting edge technology. How big does one get is a whole new argument. However, the widespread opening of overseas branches of the US companies in China, made IP transfer a bit reckless. This no doubt would have come as a shocker for the owners as there is a danger for new competitors to sprout using the stolen technology leading to billions dollars loss.

This is a serious issue since China has been consistently notorious overseas when it comes to IP theft. Here are a few that got the spotlight as listed by Jeff Ferry, CPA Research Director:

  • In 2004 Cisco Systems, took Huawei to court for stealing its core router software code and using it in Huawei routers. Huawei routers, widely used in China and Europe, have played a key role in Huawei’s growth into a $95 billion global telecom equipment giant
  • In 2011, AMSC filed the largest-ever IP theft case in a Chinese court, seeking $1.2 billion in compensation for their losses. AMSC partnered with a Chinese maker of the wind turbine hardware, Sinovel, to sell into the Chinese market. AMSC sales rose rapidly into the hundreds of millions of dollars. In 2011, AMSC discovered that Sinovel had an illegal copy of the entire AMSC software code on one of their windmills
  • In 2015, the federal government charged six Chinese citizens with stealing wireless communications technology from two Silicon Valley microchip makers, Avago and Skyworks, and launching their own company to sell that technology in China. 

Apart from this, Huawei has also been accused of stealing the patented smartphone camera technologies a couple of months ago.

The biggest concern for the developers regarding patenting can be distilled down to two words — infinite leverage.

They fear that the aspirants will either be squeezed midway or get discouraged altogether from accessing state of the art technology, which again would lead to outcomes like the much dreaded AI winter.

More Great AIM Stories

Ram Sagar
I have a master's degree in Robotics and I write about machine learning advancements.

Our Upcoming Events

Conference, Virtual
Genpact Analytics Career Day
3rd Sep

Conference, in-person (Bangalore)
Cypher 2022
21-23rd Sep

Conference, in-person (Bangalore)
Machine Learning Developers Summit (MLDS) 2023
19-20th Jan, 2023

Conference, in-person (Bangalore)
Data Engineering Summit (DES) 2023
21st Apr, 2023

Conference, in-person (Bangalore)
MachineCon 2023
23rd Jun, 2023

3 Ways to Join our Community

Discord Server

Stay Connected with a larger ecosystem of data science and ML Professionals

Telegram Channel

Discover special offers, top stories, upcoming events, and more.

Subscribe to our newsletter

Get the latest updates from AIM

Council Post: How to Evolve with Changing Workforce

The demand for digital roles is growing rapidly, and scouting for talent is becoming more and more difficult. If organisations do not change their ways to adapt and alter their strategy, it could have a significant business impact.

All Tech Giants: On your Mark, Get Set – Slow!

In September 2021, the FTC published a report on M&As of five top companies in the US that have escaped the antitrust laws. These were Alphabet/Google, Amazon, Apple, Facebook, and Microsoft.

The Digital Transformation Journey of Vedanta

In the current digital ecosystem, the evolving technologies can be seen both as an opportunity to gain new insights as well as a disruption by others, says Vineet Jaiswal, chief digital and technology officer at Vedanta Resources Limited

BlenderBot — Public, Yet Not Too Public

As a footnote, Meta cites access will be granted to academic researchers and people affiliated to government organisations, civil society groups, academia and global industry research labs.