Видео

You suck

Most likely.....

@@user-bg2xn8wd9y It must be Akida. No other SNN processor comes close.

If you look at one shot in video of computer screen you can see the words bcdemo. Brainchip demo

2:42

Hi there, could you drop a link to the paper & go into what exactly you would like explained? An explanation of the complete paper might be beyond the scope of a youtube comment.

@@FraunhoferHHI Concept Relevance Propagation part sir

That’s a profound theory. Thank you for sharing.

It’s not a regression if it’s a new way to use light waves for data communications. We currently use RF for cell phones and AM/FM, TV, Wi-Fi. So why not use light, if at all possible?

Hi, thanks for your thoughts. The concept may be known but the technology is brand new. The high requirements for mobile communication, especially the next generation 6G, make disruptive technologies that add new spectrum to the already overcrowded RF band necessary. HD television was also invented earlier but the great success came much later.

Well, not exactly. Whoever had the chance to experience IrDA and LiFi in real life will immediately see the significant differences. LiFi offers full mobility, data rates up to multiple gigabit per seconds and is using state of the art technology. With a single LiFi access point you can cover several square meter of a room at several meters distance.

Lifi uses infrared light, so human eyes do not see it. I.e. it works in the dark.

It transmits on infrared wavelengths that are invisible to the human eye.

@@littlbilly76 LiFi standard compliant device operates from 380 nm to 780 nm wavelength region. This devices operate on one or multiple visible light frequency bands.

Hi, thanks for your question! LiFi can also use wavelength in the near infrared region. A lot of optical components are available at 850 nm or 940 nm, which is not visible for the human eye. Therefore the LiFi systems can also be operated in the dark, without distracting the user of the system.

@@FraunhoferHHI thanks for clarifying

Ye, and ordinary curtains do not block light, some video sensors can easily see light through fabric curtains or any other thin material as if its not there. So hackers would be able to steal data through windows from very afar, kilometers away even. If hackers would have point laser emitters, they may even interact with Li-Fi network in chosen apartment and hack into the system and control it.

@Hokunin The same concern applies with wifi, but we don't have this problem because encryption exists. Lifi would probably just do the same thing @SMorrisRose I imagine it wouldn't be too hard to have windows with special glass or coating that attenuates the lifi wavelength (e.g. infra-red at 700nm). This all also ignores the fact that the attenuation over distance is probably far higher than radio signals.

@@fbcpck ye, the thing is, it would be awfully suspicious for some dude to walk around with laptop and other hardware from floor to floor sitting on the stairs and then walking with all that to another apartment building, looking for weakly defended wifi networks, cuz each wifi network doesn't extend too far, signal becomes weak even at the next room of one living apartment. Strangers also won't be allowed to enter any appartment building. So mostly they don't bother with wifi. Whereas with Li-Wi, hackers can have any sized equipment at hand, no need to move around, and they will have thousands of windows available for probing, just keep clicking mouse button. It will become physically easier, less bothersome, so many would be hackers would start trying to hack.

Hi, thanks for your input! It is true that light penetrates windows but optical communication cells can be sharply confined. Cell sizes of only a few square meter are possible, giving you the control were your data goes. This is simply not possible with RF. Thereby LiFi can add another layer of security to wireless communication, next to all other commonly used encryption methods. Moreover it is robust against jamming, which is an important security feature in industrial or medical environments.

Danke für Ihr Interesse! Tatsächlich gibt es neben dem Fraunhofer HHI als Technologielieferanten bereits einige Start-ups, die LiFi Systeme kommerziell anbieten. Für uns als Fraunhofer HHI liegt der Fokus nicht nur darauf die Technologie in der Anwendung zu demonstrieren, sondern auch stetig weiterzuentwickeln.

Wenn das Lichtsignal lediglich für einen kurzen Augenblick unterbrochen wird, werden die Daten gepuffert und die Übertragung kann im Anschluss so gut wie unbemerkt fortgesetzt werden. Ansonsten wird die Datenübertragung pausiert. Bei vielen Anwendungen wird LiFi als Entlastung für die bestehenden funkbasierten Systeme oder als Datenhotspot mit sehr hohen Geschwindigkeiten eingesetzt. Dort kann dann das Funksystem für den Moment übernehmen. Wenn das Lichtsignal dann wieder Verbindung aufnehmen kann, steht wieder die volle Geschwindigkeit per Licht zur Verfügung.

Mach mal!

AV1 is supposed to be open. This is the genius of FOSS. It enables faster improvements. The downside in this case is it's a lot harder for hardware to keep up. Maybe if AMD made their compute cores easier to access, everyone with a GPU could have it accelerated even without specific HW support.

There are already publicly available VVC integrations, e.g.: github.com/MartinEesmaa/VVCEasy Integration into the main repositories of these projects is ongoing so stay tuned. More info on VVC deployment: jvet-experts.org/doc_end_user/documents/27_Teleconference/wg11/JVET-AA0021-v1.zip

Are manufacturers releasing TV's with GPU's installed? That would give quite an increase in cost. Professional hardware and software might go that route but for consumer or professional display systems dedicated ic's (encoding / decoding hardware) are used. Professional and some home video editing coding software provides the option for a render to be done via the CPU or GPU. Dedicated ic's would be optimised to do the one task within the timeframe needed keeping the cost down.

It's better than AV1

Arun, 10 seconds of Googling would show you av1 is far inferior.

It's like passing the speaking rock around, no rock, no speak.

Sure it can, 3GPP evaluated exactly these codecs for different 5G video scenarios together with the proponents of these codecs: www.3gpp.org/ftp/Specs/archive/26_series/26.955/26955-200.zip For Scenario 1 (Full HD) bit-rate savings over HEVC are: VVC - 36% SDR / 40% HDR EVC - 18% SDR / 21% HDR AV1 - 19% SDR / 15% HDR For Scenario 2 (4K-TV) bit-rate savings over HEVC are: VVC - 37% SDR / 38% HDR EVC - 20% SDR / 23% HDR AV1 - 18% SDR / 17% HDR

The codec is available globally and the Indian MX Player was actually the first OTT to deploy H.266 (press release from June 16th)! Find more info here: 8kassociation.com/first-usage-of-vvc-in-live-service-in-india/ And here: bestmediainfo.com/2021/06/mx-player-becomes-first-ott-to-deploy-h-266-cuts-down-video-streaming-data-consumption-into-half

We are still working on a subtiteled version. But it will go live in a few days - so stay tuned & thank you for your patience!

Dankeschön für dein Interesse - das freut uns!

We’ve been working hard on getting the encoding times down with VVC. Check out our progress on the VVenC project. It’s definitely worth a try, we are able to match encoding times of some x265 presets at much lower bitrates. www.hhi.fraunhofer.de/en/departments/vca/technologies-and-solutions/h266-vvc/fraunhofer-versatile-video-encoder-vvenc.html

Current bandwidth is not great for 4K streaming and as you know High Resolution is the future of video streaming and therefore we need a compression technique that will efficient for such resolutions... That's why we need this technology... Talking about CPU usage Don't you think device that is capable of 4K streaming must have great processing power

it was the same for both, and it's not pixelation. Those are encoding blocks.

The artifacts mostly come from RUclips transcoding. Perks of showing video comparisons on RUclips during corona. Hopefully we’ll be able to present the quality improvements without transcoding soon.