AI starts to reason like humans

relational reasoning AI
AI research

AI Relation networks

AI is extremely good at identifying and using patterns. However, when it comes to relational reasoning, artificial intelligence systems badly fail at it.

Relational reasoning is an essential part of higher thought and logic. It allows humans to make connections and compare places, sequences, entities and other variables.

Researchers at Google’s DeepMind have tackled this problem and developed a simple relational reasoning algorithm to help machines use logic. The new system relies on AI neural networks that mimic the way neurons are connected. More specifically, dedicated neural networks called relation networks work together to identify the connections and relations between entities.

Relation networks

Relation networks are AI architectures that focus explicitly on relational reasoning. The network compares pair of objects individually in order to identify the relationships that exist between the objects.

This is the first major step in building smarter AI systems capable of reasoning about the relations between entities and their properties. Unlike other reasoning networks, RNs are simple, plug-and-play, and flexible, allowing researchers to use them in various fields.

Researchers used RN-augmented networks to a variety of tasks that heavily rely on relational reasoning, including visual QA, text-based QA, and dynamic physical systems. Thanks to the versatility of these networks, researchers were able to achieve state-of-the-art, super-human performance.

The visual QA test

Visual QA tests challenge AI systems to answer questions about an image. This task requires high-level scene understanding, and complex relational reasoning – spatial and otherwise. To make this task even more difficult, oftentimes vast knowledge of the world is not available in the training data. Moreover, questions contain ambiguities and linguistic biases.

DeepMind researchers used two versions of the CLEVR dataset for this particular test: the pixel version, in which images were represented in standard 2D pixel form, and a state description version, where images were explicitly represented by state description matrices with factored object descriptions.

Here are the results:

Our model achieved state-of-the-art performance on CLEVR at 95.5%, exceeding the best model trained only on the pixel images and questions at the time of the dataset’s publication by 27%, and surpassing human performance in the task (see Table 1 and Figure 3). These results – in particular, those obtained in the compare attribute and count categories – are a testament to the ability of our model to do relational reasoning. In fact, it is in these categories that state-of-the-art models struggle most. Furthermore, the relative simplicity of the network components used in our model suggests that the difficulty of the CLEVR task lies in its relational reasoning demands, not on the language or the visual processing.

Using relation networks to improve AI reasoning

Relation networks are very useful for identifying inter-entity relations. By embedding relation networks modules into deep learning architectures, researchers can improve  AI performance on tasks that involve relational reasoning.

Experiments revealed that RNs are very flexible and are capable of structured reasoning even with unstructured inputs and outputs.

Relation networks can be used to enhance reinforcement learning  agents’ capacity to understand rich scenes, model social networks, and solve abstract problems.

For more information, you can check out the research paper here.

YOU MAY ALSO LIKE: Chinese companies may soon overtake Google in AI research

Follow The AI Center on social media:

Maddie Blau

I strongly believe that Artificial Intelligence is the future of technology. AI research has yielded significant advancements in recent years and this is only the beginning.

Join me as I track the latest progress in AI research.
AI stock exchange
AI in real life
Canada’s first AI exchange-traded fund enters the market

  Canada launched the first artificial intelligence exchange-traded fund with the ticker MIND. The Horizons Active A.I. Global Equity ETF is an investment strategy run by an AI system. The AI algorithm analyzes financial data to extract patterns in order to determine what are the best investments to make. MIND …

smartest AI
AI research
Are you curious to see who’s the smartest AI in the world?

  The AI world is dominated by fierce competition. Well-known tech giants such as Google, Microsoft, Intel, NVIDIA and other are working at full speed to push AI research forward and innovate the field. The question is: who’s the smartest AI in the world? This is a very difficult question …

google assistant natural sound voice
AI research
Google Assistant’s voice is now more realistic thanks to WaveNet

Getting speech synthesis algorithms to sound natural is a hard nut to crack. AI researchers have been struggling to make digital assistants sound more natural for years, and the results are indeed impressive. However, there’s still something that doesn’t sound right and users can always tell if they’re talking to …