Reconstruction of neural circuitry at single-synapse resolution is an attractive target for improving understanding of the nervous system in health and disease. Serial section transmission electron microscopy (ssTEM) is among the most prolific imaging methods employed in pursuit of such reconstructions. We demonstrate how Flood-Filling Networks (FFNs) can be used to computationally segment a forty-teravoxel whole-brain Drosophila ssTEM volume. To compensate for data irregularities and imperfect global alignment, FFNs were combined with procedures that locally re-align serial sections and dynamically adjust image content. The proposed approach produced a largely merger-free segmentation of the entire ssTEM Drosophila brain, which we make freely available. As compared to manual tracing using an efficient skeletonization strategy, the segmentation enabled circuit reconstruction and analysis workflows that were an order of magnitude faster. Read More
Monthly Archives: August 2019
Artificial intelligence predicts which movies will succeed—and fail—simply from plot summaries
Artificial intelligence (AI) still can’t see the future, but a new algorithm may come close: Using nothing but written movie summaries, the AI can consistently tell which films will play well—or rottenly—to critics and audiences. If the model can be further refined, it could one day help producers predict whether a movie will be a flop at the box office, before it’s even made. Read More
Advances in Conversational AI
Dialogue research is a crucial component of building the next generation of intelligent agents. While there’s been progress with chatbots in single-domain dialogue, agents today are far from capable of carrying an open-domain conversation across a multitude of topics. Agents that can chat with humans in the way that people talk to each other will be easier and more enjoyable to use in our day-to-day lives — going beyond simple tasks like playing a song or booking an appointment.
Generating coherent and engaging responses in conversations requires a range of nuanced conversational skills, including language understanding and reasoning. Facebook AI has made scientific progress in dialogue research that is, in the long run, fundamental to building more engaging, personable AI systems. In this blog post, we describe new open source data sets, algorithms, and models that improve five common weaknesses of open-domain chatbots today: consistency, specificity, empathy, knowledgeability, and multimodal understanding. Read More
Every step you take
Kashgar, China — You expect to have problems when reporting on sensitive stories in China. People following you, thugs blocking your way, obstacles springing up in the most innocuous of places — it shocks you at first but eventually you get used to things like that. It comes with the territory.
But when we went to the region of Xinjiang a few months back, it was a whole other level — surreal, sinister and sometimes comical all rolled into one. Read More
Watch Hong Kong Protesters Use Lasers To Disrupt Facial Recognition Cameras
Protesters in Hong Kong appear to be keenly aware of the Chinese mainland’s all-pervading Orwellian surveillance system and facial recognition software, and they are already taking action to thwart such systems installed in Hong Kong.
The UK Independent reports based on a viral video posted by a freelance journalist this week:
Protesters in Hong Kong are using lasers to blind security forces and avoid facial recognition cameras used by authorities.
Hong Kong protestors are on another level. Here they’re using lasers to avoid facial recognition cameras. A cyber war against Chinese artificial intelligence. pic.twitter.com/t1hIczr5Go — Alessandra (@alessabocchi) July 31, 2019
Read More
AI translation advances are now posing a risk to jobs
Researchers from MIT, along with Google’s AI labs, recently announced that they have developed an AI translation tool capable of translating ancient languages.
Using a machine learning system, the tool was able to successfully translate Linear B, the first known example of the Greek language.
Neural machine translation (NMT) technology has revolutionised translation in recent years, with the capabilities of AI-based translation rapidly improving and tech giants such as Google, Amazon and Microsoft investing in this area. Read More
Brain Talker Makes “Mind Reading” Possible—Tianjin Creates the World’s First Brain-Computer Codec Chip
The world’s first Brain-Computer Codec Chip (BC3), Brain Talker, was announced on May 17, 2019, during the 3rd World Intelligence Congress at Tianjin. The Brain Talker was a joint effort of Tianjin University and China Electronics Corporation with fully independent intellectual property.
This BC3 chip was specially designed to improve the Brain-Computer Interface (BCI) technology, which aims at decoding a user’s mental intent solely through neural electrical signals, without the use of the human body’s natural neuromuscular pathways. Read More
Haier, China Mobile and Huawei Launch the World's First AI+5G Interconnected Factory
Haier, China Mobile, and Huawei jointly launched the world’s first AI+5G interconnected factory at the 2019 World Industrial Internet Conference (WIIC) in Qingdao. This improvement refines future smart manufacturing by innovating and transforming enterprise organizations, business models, and ICT technologies, and integrating key technologies such as AI and 5G.
China Mobile Shandong and Huawei helped complete 5G base station deployment for Haier’s interconnected factory in Zhongde Industrial Park, Huangdao District, Qingdao on June 13, 2019. Read More
The AI Podcast, Ep. 32: Deep Learning Pioneer Andrew Ng on AI as the New Electricity
Purple shirts, haircuts, and cats. How are these three all related? According to deep learning pioneer Andrew Ng, they all played a part in AI’s growing presence in our lives. Ng, formerly of Google and Baidu, and the founder of his new company, Deeplearning.ai, joined this week’s episode of the AI Podcast to share his thoughts on AI being the new electricity. Read More
Developing Brain Atlas through Deep Learning
Neuroscientists have devoted significant effort into the creation of standard brain reference atlases for high-throughput registration of anatomical regions of interest. However, variability in brain size and form across individuals poses a significant challenge for such reference atlases. To overcome these limitations, we introduce a fully automated deep neural network-based method (SeBRe) for registration through Segmenting Brain Regions of interest with minimal human supervision. We demonstrate the validity of our method on brain images from different mouse developmental time points, across a range of neuronal markers and imaging modalities. We further assess the performance of our method on images from MR-scanned human brains. Our registration method can accelerate brain-wide exploration of region-specific changes in brain development and, by simply segmenting brain regions of interest for high-throughput brain-wide analysis, provides an alternative to existing complex brain registration techniques. Read More
Rick's Cafe AI 