BRN Discussion Ongoing
- Thread starter TechGirl
- Start date
Frangipani
Top 20
C'mon TCS....ya know you want Akida 2.0 and TENNs.......sign here pls
Am I the only one who thinks that it is not going to be consultancy companies such as TCS or Accenture who are potentially going to sign an IP license with us, but rather their customers?
CHIPS
Regular
Tata consultancy services with gesture recognition
What if gesture control could work smoothly, without being limited by hardware? | Tata Consultancy Services - Research
What if gesture control could work smoothly, without being limited by hardware? Our researchers have patented a low-power, ultrasound-based gesture detection system that leverages MIMO setups and CIR estimation for improved accuracy. By converting traditional neural networks into spiking neural...www.linkedin.com
Didn't BrainChip withdraw Akida 2 in favour of some company's development last year or so? Now they are presenting Akida 2 in Nuremberg and, at the same time, Tata informs about their new patent. Coincidence?
SERA2g
Founding Member
Found your post searching for Arquimea on Tsex.Wonder if we'll get an intro to these guys as some point. Co funded by the EU and have also participated in space programs with ESA and NASA.
Like to think someone at ESA may discuss us with them sometime?
Exploring the future of the Space sector
In this article we explore the challenges the space sector will face in the coming years, as well as the trends, direction and future of the space sector.www.arquimea.com
Spanish technology for the European XFEL, the world's most important scientific X-ray facility
Neuromorphic vision is a field that draws on the workings of the human visual system to develop electronic systems that process visual information efficiently and in real time.
What are the challenges of implementing neuromorphic vision in AI?
Trends
December 23, 2024
Neuromorphic vision is a field that draws on the workings of the human visual system to develop electronic systems that process visual information efficiently and in real time. This approach uses sensors and algorithms designed to mimic the biological properties of the eye and brain.
Instead of capturing data in fixed frames like traditional cameras, neuromorphic sensors record individual events (changes in light intensity) at each pixel. This makes them highly efficient in terms of energy consumption and processing speed.
Origins of Neuromorphic Vision
The term ‘neuromorphic’ was coined by Carver Mead in the 1980s. Mead, a pioneer in microelectronics, proposed to design electronic systems inspired by the structure and function of the human brain. Since then, research in neuromorphic sensors has evolved, with key milestones such as the development of event cameras (e.g. Dynamic Vision Sensor, DVS) that mimic the behaviour of the human eye.
Relationship with Artificial Intelligence (AI)
Neuromorphic vision is closely linked to AI, providing highly optimised and relevant data for the training and execution of deep learning and machine learning algorithms. Some of its main contributions are:
- Real-time processing: data obtained from neuromorphic sensors allows AI models to react immediately, useful in applications such as autonomous driving and robotics.
- Energy efficiency: neuromorphic vision significantly reduces energy consumption compared to traditional cameras, improving the sustainability of AI-based applications.
- Non-redundant data: event-specific detection allows AI systems to work with non-redundant data, improving accuracy in tasks such as object recognition or navigation.
Current and Future Impact
The implementation of neuromorphic vision is highly recommended in sectors where low latency and energy efficiency are essential and robust real-time event processing is required. Therefore, neuromorphic vision has promising applications in sectors such as:
– Robotics: enabling improved visual perception of robots for navigation and manipulation in complex environments.
– Autonomous driving: enables fast and efficient detection of both objects and obstacles.
– Medical devices: supports some technologies such as visual prostheses or biomedical analysis.
– Security and surveillance: provides highly accurate real-time detection of suspicious movements and critical events.
– Industry and automation: aiding quality inspection systems, tracking objects on assembly lines, and industrial IoT systems.
The combination of neuromorphic vision and AI will transform the way machines perceive and understand the environment, bringing them closer to human biological processing.
What are the challenges of implementing neuromorphic vision in AI?
Implementing neuromorphic vision in artificial intelligence presents several challenges, whether technical, economic or practical:
– Development of specialised hardware: neuromorphic sensors require advanced chips that mimic the neural activity of the brain, which are expensive and technically complex to manufacture.
– Unconventional data processing: instead of conventional images, neuromorphic sensors generate data in event format, which requires specific algorithms and new paradigms to interpret the data.
– Specialised learning algorithms: New algorithms, such as spiking neural networks (SNNs), are needed that are compatible with the asynchronous and event-driven nature of the data.
– Scalability: Algorithms and systems need to be scalable for large-scale applications, which has not yet been fully achieved.
– Lack of expertise and training: There are few experts in neuromorphic vision, and it takes time and resources to build technical teams.
In summary, neuromorphic vision has transformative potential in multiple industries. Its implementation is strategic for companies seeking technological advantage in artificial intelligence applications.
Neuromorphic Vision and Artificial Intelligence in ARQUIMEA
ARQUIMEA, from its research center located in the Canary Islands, has a research orbital dedicated to robotics and another to Artificial Intelligence that develops projects that explore the potential of neuromorphic vision.
In addition, all ARQUIMEA Research Center projects belong to the QCIRCLE project, co-funded by the European Union, which aims to create a center of scientific excellence in Spain
An Embedded Software Research Engineer from Arquimea just liked one of Brainchip's recent LinkedIn posts.
Here's the post and like:
https://www.linkedin.com/posts/brai...p&rcm=ACoAABNRhjYBFfSUxFCyQN7T30eFUWBJTh-LcmM
Nothing concrete but thought you'd like to know! Arquimea could be a rabbit hole worth chasing given we now know someone working there is aware of Brainchip.
Frangipani
Top 20
#EmbeddedWorld2025 marked another exciting milestone for BrainChip. | BrainChip
#EmbeddedWorld2025 marked another exciting milestone for BrainChip. Over the course of the event, we had the opportunity to engage with developers, partners, and industry leaders from around the globe, all sharing a vision for more intelligent, efficient, and responsive edge AI. Our booth was...
View attachment 79220
The two gentlemen standing next to Alf Kuchenbuch and Gilles Bézard in the top picture are Jules Lecomte and Axel von Arnim from fortiss, by the way!
Great to see them visiting the BrainChip booth!
Coincidentally, I had posted another picture with Jules Lecomte in it just two hours ago, regarding the CORINNE project… Or was it telepathy?
Anyway, today was definitely not Jules Lecomte’s first encounter with Akida:
https://thestockexchange.com.au/threads/brn-discussion-ongoing.1/post-447484
View attachment 79221
(…)
View attachment 79222
Ladies and gentlemen, we have a new partner - fortiss!
Fortiss is the München/Munich-based Research Institute of the Free State of Bavaria for software-intensive systems (https://www.fortiss.org/en/about-fortiss).
It is a non-profit limited company, whose shareholders are the Free State of Bavaria (2/3) and the Fraunhofer-Gesellschaft / Fraunhofer Society (1/3).
They have been doing neuromorphic research for years, collaborating extensively with Intel, and as partners in the Human Brain Project also experimented quite a bit with SpiNNaker. In addition, they have also worked with IBM:
https://thestockexchange.com.au/threads/brn-discussion-ongoing.1/post-418482
Neuromorphic Computing - fortiss
The fields of competence at the fortiss Institute face the current trends in the development of software and its use in practice within an ambitious research program.
As I mentioned in my above post last night, fortiss and neuroTUM (TU München) jointly organised a Neuromorphic Hackathon in mid-November, where a student team mentored by Jules Lecomte (fortiss), Gregor Lenz (Neurobus) and Arunkumar Rathinam (University of Luxembourg) won the challenge using Akida.
Note, though, that fortiss have also newly partnered with Innatera alongside us.
(brochure in German only)
Research partnerships:
Last edited:
Yep. When I heard the BrainChip team proudly mention that existing partners visited their booth I kinda of thought surely they already communicate with each other enough. And don’t we already have enough customers to focus on. I think it’s fair enough having Tony as a keynote speaker, I can see the value. But how much value are we getting from our booth that’s 1/10 tenth of the size of Edge Impulses.
HopalongPetrovski
I'm Spartacus!
Hi GazDix.
I sympathise with what you're expressing and having worked at a company that went the conference and trade show route, I felt the same way at the time.
And it took a hell of a lot of persistence and years of apparently wasted expense, before it eventually paid off.
In many cases the opportunity tide turned on the established reputation of being a major "player" in the market and getting to know and be known by the other significant entity's and their key personal.
It's the old "not what you know but who" that often opens doors to partnerships and opportunities that don't otherwise get advertised and much of it is done on a nod and a handshake, amongst known, and more significantly, trusted colleagues.
I could never keep up with him, but my boss who drove this strategy would spend each night of the trade shows out drinking and carousing with his contemporaries until they eventually became pals. Pals who do deals with each other.
It's also a good place to meet new players who sometimes have access to innovative or cheaper options that can be exploited.
It's still cut throat business, but eventually just showing up and being seen amongst the big boys starts to lend one some gravitas.
Of course you still need a compelling and competitive offering but waiting around to be discovered can take significantly longer, and faster growth is achieved through perceived presence in the market.
After initially being sceptical, then seeing it play out over time, I came away a believer.
So much in business comes down to "who" you know, and "who" knows you.
It was actually what first drew me to Sean.
His contacts. But, like I said, it takes time.
And longer than many of us wagered for, I think.
Although also still think that given they are now pushing for the redomicile, they must be close to a significant breakthrough.
I certainly bloody hope so.
The drinking part sounds appealing.Hi GazDix.
I sympathise with what you're expressing and having worked at a company that went the conference and trade show route, I felt the same way at the time.
And it took a hell of a lot of persistence and years of apparently wasted expense, before it eventually paid off.
In many cases the opportunity tide turned on the established reputation of being a major "player" in the market and getting to know and be known by the other significant entity's and their key personal.
It's the old "not what you know but who" that often opens doors to partnerships and opportunities that don't otherwise get advertised and much of it is done on a nod and a handshake, amongst known, and more significantly, trusted colleagues.
I could never keep up with him, but my boss who drove this strategy would spend each night of the trade shows out drinking and carousing with his contemporaries until they eventually became pals. Pals who do deals with each other.
It's also a good place to meet new players who sometimes have access to innovative or cheaper options that can be exploited.
It's still cut throat business, but eventually just showing up and being seen amongst the big boys starts to lend one some gravitas.
Of course you still need a compelling and competitive offering but waiting around to be discovered can take significantly longer, and faster growth is achieved through perceived presence in the market.
After initially being sceptical, then seeing it play out over time, I came away a believer.
So much in business comes down to "who" you know, and "who" knows you.
It was actually what first drew me to Sean.
His contacts. But, like I said, it takes time.
And longer than many of us wagered for, I think.
Although also still think that given they are now pushing for the redomicile, they must be close to a significant breakthrough.
I certainly bloody hope so.
Bombersfan
Regular
Yeah not attending world renowned tech shows that a lot of big players attend to keep showcasing a revolutionary new product, and avoiding face to face confrontation at agm’s would be a real show of strength. Traction is needed soon, but what a bullshit couple of comments.
Stable Genius
Regular
DingoBorat
Slim
"Ladies and gentlemen, we have a new partner - fortiss!"Ladies and gentlemen, we have a new partner - fortiss!
Fortiss is the München/Munich-based Research Institute of the Free State of Bavaria for software-intensive systems (https://www.fortiss.org/en/about-fortiss).
It is a non-profit limited company, whose shareholders are the Free State of Bavaria (2/3) and the Fraunhofer-Gesellschaft / Fraunhofer Society (1/3).
They have been doing neuromorphic research for years, collaborating extensively with IBM and Intel.
Neuromorphic Computing - fortiss
The fields of competence at the fortiss Institute face the current trends in the development of software and its use in practice within an ambitious research program.www.fortiss.org
As I mentioned in my above post last night, fortiss and neuroTUM (TU München) jointly organised a Neuromorphic Hackathon in mid-November, where a student team mentored by Jules Lecomte (fortiss), Gregor Lenz (Neurobus) and Arunkumar Rathinam (University of Luxembourg) won the challenge using Akida.
Note, though, that fortiss have also newly partnered with Innatera alongside us.
View attachment 79266
View attachment 79267
(brochure in German only)
Research partnerships:
View attachment 79268 View attachment 79269
...It is a non-profit limited company..
"It says PROFIT! I'm gonna be rich!"
Thanks Frangipani, Tata applied for the patent on 14/12/22 so they have been playing with it for over 2 years.
We just might be getting closer to another validation event via Tata sometime this year.
Not sure what product use or uses they have in mind for it.
TATA did say health and Industrial.
Now that Onsor has the honor of the 1st wearable i think we see plenty of work being done with AKIDA on all sorts of afflictions.
I was thinking the same thing, We have had enough exposure to gain interest but obviously The Team are still chasing.......
GazDix
Regular
Glad the Hawks will smash you tonight sunshine.
Bombersfan
Regular
Ouch that’s a zinger.
I wonder if Pepsi will stop spending millions on Super Bowl ads and maccas will stop sponsoring almost everything because they already have enough customers. I’m sorry but that was the dumbest comment I’ve heard in a while. Let’s not showcase the astronomical developments with LLM’s and hope people come to us and ask us “whats new”?
Found this.. not Brainchip related but interesting
www.linkedin.com
jp.ubuntu.com
Arm and intel partner
canonical.com
Ubuntu | Canonical
従来の大規模言語モデル（#LLM ）の重大な欠点の1つは、すでにモデル内にある静的な情報に依存することです。基本的に、プロンプトを理解して応答するのは得意かもしれませんが、多くの場合、最新の情報や特殊性の高い情報は提供できません。そこで活躍するのが検索拡張生成（RAG）です。RAGは、現在の情報と新しい情報を組み込み、これらのモデルの信頼できる情報源とすることで、この欠点を克服します。 あらゆるハードウェアやクラウドで動作する最高レベルの #生成AI ツールを学習して使用しましょう👉
Ubuntu
Ubuntuは、デスクトップPCやクラウド、インターネットに接続されたあらゆる機器まで、すべての環境において動作可能なオープンソースのソフトウェアオペレーティングシステムです。
jp.ubuntu.com
Arm and intel partner
Silicon Programme | Partners | Canonical
Canonical and our silicon partners have provided market-leading solutions for customers seeking alternatives to complex and costly proprietary operating environments.
canonical.com
Quiltman
Regular
Similar threads
