I was at the talk, and that's pretty much exactly what they had done. There weren't any detailed tech specs, but the essence was that they had simply put a lot of big discs in a box the size of a fridge, added as many tuner cards as there are multiplexes (6?), and built some navigation software on top.
My experience in European trading floors mirrors this. The environments were designed for flexibility and to put the analytics in the hands of the traders. At the first bank I worked for, the entire analytic library could be called from Excel and Excel VBA along with C++ etc. Realtime data could bound from reuters, Bloomberg, and the various internal publish/subscribe mechanisms.
The traders ran their own sheets based on these tools which constantly evolved. My first job there was to make a prototype production application of a model one of the traders had built in Excel. To be honest, the money would have been better spent if I had tidied up the existing Excel - at least when the model changed the following week the guy could have made his own updates.
The bottom line is that *everyone* in finance knows Excel. It is the de facto standard tool for front end data manipulation. Hence it is a natural choice for rapid development of data minipulation applications. You could suggest that they use another tool instead, but it would be like suggesting that everyone start using base 12 arithmetic - it might have some advantages, but everyone already knows base 10.
Hi. I would be very interested if someone could suggest or reccommend software that functions like MythTV or any of the products mentioned in the summary above, but which does not have TV functionality and instead focusses on music.
I have a miniITX board that I would like to use as a music server. Audio out would go through to my HiFi. Video out would be to my TV. control would be through remote control.
Things I want on a phone: 1. Small 2. Lots of space for contacts 3. Synch with Outlook 4. Some flash memory with a USB socket, like a USB memory stick 5. Well designed UI 6. Good audio quality 7. Shold look recognisably like a phone 8. Predictive text
Things I don't want on a phone: 1. Camera 2. Video camera 3. Games 4. Audio recorder 5. mini qwerty keyboard 6. flashlight 7. GPS 8. Compass 9. Microsoft Office 10. A meda player
Things that are acceptable as long as they don't get in the way:
1. GPRS 2. Some kind of WAP/internet thing 3. Bluetooth 4. a Java runtime
Incidentally, I had the same phone as you until it broke. subsequent models have been larger and less easy to use.
Also, with reference to "must look like a phone", when Nokia released the 6230 last year, almost everyone I know bought one within 2 months. This was because it didn't look like it was designed by a 12 year old like the previous two years output.
Can you give me an example of a large scale terminal services installation, using dedicated diskless clients? I have only ever seen microsoft's implementation of RDP used for remote system administration, rather than true X-like thin clients.
On the other hand, I have seen a number of large Cytrix MetaFrame installations, especially at hospitals - most likely this is what you have seen.
So just to be clear, Microsoft is certainly not leading this market, though they do have some strong assets in place.
Really? You'd guess that there's global interest in this? Google news reckons that there's 2570 articles from around the world that it knows about. Who do you think those people are rooting for? (ducks...)
A second factor is that these statistical speech recognition systems require extensive data for their language model. Building such a system requires recording real speech, segmenting it and creating a set of examples from which to compute the probabilities, which requires some knowledge of acoustic phonetics, and doing the computation for the model.
A Language model is typically built from available text sources such as newspaper text. It does not require that you have recorded speech. Generating an acoustic model, on the other hand, does require accurately transcribed, recorded speech, and lots of it.
A lot of a person's speech recognition ability comes from context
That is also true for a reco system. Typically a Large vocabulary continuous reco system will use a tri-gram language model in order to take into account contextual information. This means that likelyhood of the candidate words is looked up based on the previous three words, which has been shown to provide a sufficient degree of contextual information to distinguish between most commonly encountered homophones.
It won't. Speech recognition will never be big on the desktop except as an accessibility feature or for niche dictation circumstances such as medical examination or autopsy. (Incidentally, the recognition accuracy for these special applications is better than standard text due to the prevalence of long, obscure words - think legalese). The fundamental reason for this is that where a keyboard and mouse is available, they offer a better control method than speech.
Speech reco *will* become pervasive in applications where it is not possible to use a standard keyboard/mouseinterface, for example telephone services or on small form factor devices such as telephones and PDAs. However no one is going to make a great deal of money on this.
It wasn't. IBM had been doing theoretical work in speech recognition for some time, and had developed their own speech engine and acoustic and language models for supported languages. Sooner or later thay might even get around to doing something commercial with the technology!
IBM also has (or rather had in 98,99,2000) a grammar based recognition system based on the same engine, but using compiled grammars and naturally a cut down acoustic model dependant on the contents of the grammar. There was also a toolset, supporting compiling grammars from BNF, building speech telephony applications and so forth.
IBM Hursley labs had a name dialler 5 years ago that let you phone the computer, say the name fo the person you wanted to speak with, and get put through. They also had a system that provided weather forecasts based on the name of the city or country you said. I was pleased to name the latter "Global Weather Information System" or GWIS, pronounced Gee-whizz. Both ran on the machine under my desk. Both worked reasonably well, especially given that a lot of the acoustic models for names and places were automagically generated.
1. get one of those foot rest things that lets you support your feet in a variety of different angles. I found that they made quite a difference to my posture (in an aeron, which I generally like)
2. Take up some form of exercise that strengthens your back. I had a pretty weak back after tearing a load of muscles up one side playing field hockey 6 years ago. After I started using the rowing machine (ergometer) at the gym, I found my comfort in front of a machine improved a great deal. It's worth asking the gym instructor to show you how to use it properly, since many people don't, and that seriously reduces the results.
The explanation for this is quite messy Not really. In fact it's pretty elegant and simple. But the key point is that a 44.1k DAC is most certainly up to the task of doing a 16k tone precisely because of Nyquist.
How does establishing identity help in the slightest when the stated aim is to prevent another plane-as-suicide-bomb attack. The answer is that it doesn't in the slightest, as many others have pointed out, and may in fact be detrimental. The only reason I can see is that it makes the travellers *feel* safer, and has side effects in reducing other types of crime such as smuggling. It does not raise any bar except that under which I must contort myself when visiting the US.
Slashdot Mirror
I think you are thinking of docomo.
I was at the talk, and that's pretty much exactly what they had done. There weren't any detailed tech specs, but the essence was that they had simply put a lot of big discs in a box the size of a fridge, added as many tuner cards as there are multiplexes (6?), and built some navigation software on top.
Quorum Tools make some interesting (Linux based) products in this space.
My experience in European trading floors mirrors this. The environments were designed for flexibility and to put the analytics in the hands of the traders. At the first bank I worked for, the entire analytic library could be called from Excel and Excel VBA along with C++ etc. Realtime data could bound from reuters, Bloomberg, and the various internal publish/subscribe mechanisms.
The traders ran their own sheets based on these tools which constantly evolved. My first job there was to make a prototype production application of a model one of the traders had built in Excel. To be honest, the money would have been better spent if I had tidied up the existing Excel - at least when the model changed the following week the guy could have made his own updates.
The bottom line is that *everyone* in finance knows Excel. It is the de facto standard tool for front end data manipulation. Hence it is a natural choice for rapid development of data minipulation applications. You could suggest that they use another tool instead, but it would be like suggesting that everyone start using base 12 arithmetic - it might have some advantages, but everyone already knows base 10.
IIRC he seemed to be pretty clued up on all sorts of things if the ucam.* groups can be trusted.
that's all.
Hi. I would be very interested if someone could suggest or reccommend software that functions like MythTV or any of the products mentioned in the summary above, but which does not have TV functionality and instead focusses on music.
I have a miniITX board that I would like to use as a music server. Audio out would go through to my HiFi. Video out would be to my TV. control would be through remote control.
any thoughts?
Things I want on a phone:
1. Small
2. Lots of space for contacts
3. Synch with Outlook
4. Some flash memory with a USB socket, like a USB memory stick
5. Well designed UI
6. Good audio quality
7. Shold look recognisably like a phone
8. Predictive text
Things I don't want on a phone:
1. Camera
2. Video camera
3. Games
4. Audio recorder
5. mini qwerty keyboard
6. flashlight
7. GPS
8. Compass
9. Microsoft Office
10. A meda player
Things that are acceptable as long as they don't get in the way:
1. GPRS
2. Some kind of WAP/internet thing
3. Bluetooth
4. a Java runtime
Incidentally, I had the same phone as you until it broke. subsequent models have been larger and less easy to use.
Also, with reference to "must look like a phone", when Nokia released the 6230 last year, almost everyone I know bought one within 2 months. This was because it didn't look like it was designed by a 12 year old like the previous two years output.
I am fully aware of that. But thanks for pointing it out.
Can you give me an example of a large scale terminal services installation, using dedicated diskless clients? I have only ever seen microsoft's implementation of RDP used for remote system administration, rather than true X-like thin clients.
On the other hand, I have seen a number of large Cytrix MetaFrame installations, especially at hospitals - most likely this is what you have seen.
So just to be clear, Microsoft is certainly not leading this market, though they do have some strong assets in place.
Ditch Kreyszig and get Glyn James - Advanced Modern engineering mathematics. Only £35 (~$65) from amazon.
or try the Cubit, which is in my opinion the prettyest available.
it's a piece of shit game
good
Because of it's title.
bad
Really? You'd guess that there's global interest in this? Google news reckons that there's 2570 articles from around the world that it knows about. Who do you think those people are rooting for? (ducks...)
They just revolved.
A second factor is that these statistical speech recognition systems require extensive data for their language model. Building such a system requires recording real speech, segmenting it and creating a set of examples from which to compute the probabilities, which requires some knowledge of acoustic phonetics, and doing the computation for the model.
A Language model is typically built from available text sources such as newspaper text. It does not require that you have recorded speech. Generating an acoustic model, on the other hand, does require accurately transcribed, recorded speech, and lots of it.
This is time-consuming.
Yes. Very.
A lot of a person's speech recognition ability comes from context
That is also true for a reco system. Typically a Large vocabulary continuous reco system will use a tri-gram language model in order to take into account contextual information. This means that likelyhood of the candidate words is looked up based on the previous three words, which has been shown to provide a sufficient degree of contextual information to distinguish between most commonly encountered homophones.
Speech reco *will* become pervasive in applications where it is not possible to use a standard keyboard/mouseinterface, for example telephone services or on small form factor devices such as telephones and PDAs. However no one is going to make a great deal of money on this.
It wasn't. IBM had been doing theoretical work in speech recognition for some time, and had developed their own speech engine and acoustic and language models for supported languages. Sooner or later thay might even get around to doing something commercial with the technology!
IBM Hursley labs had a name dialler 5 years ago that let you phone the computer, say the name fo the person you wanted to speak with, and get put through. They also had a system that provided weather forecasts based on the name of the city or country you said. I was pleased to name the latter "Global Weather Information System" or GWIS, pronounced Gee-whizz. Both ran on the machine under my desk. Both worked reasonably well, especially given that a lot of the acoustic models for names and places were automagically generated.
1. get one of those foot rest things that lets you support your feet in a variety of different angles. I found that they made quite a difference to my posture (in an aeron, which I generally like)
2. Take up some form of exercise that strengthens your back. I had a pretty weak back after tearing a load of muscles up one side playing field hockey 6 years ago. After I started using the rowing machine (ergometer) at the gym, I found my comfort in front of a machine improved a great deal. It's worth asking the gym instructor to show you how to use it properly, since many people don't, and that seriously reduces the results.
The explanation for this is quite messy
Not really. In fact it's pretty elegant and simple. But the key point is that a 44.1k DAC is most certainly up to the task of doing a 16k tone precisely because of Nyquist.
How does establishing identity help in the slightest when the stated aim is to prevent another plane-as-suicide-bomb attack. The answer is that it doesn't in the slightest, as many others have pointed out, and may in fact be detrimental. The only reason I can see is that it makes the travellers *feel* safer, and has side effects in reducing other types of crime such as smuggling. It does not raise any bar except that under which I must contort myself when visiting the US.
Thanks God somebody here knows what goodwill is.
What exactly is the reason to buy an iPod if you are not into the whole "online music store" thing?
It looks nice and has a beautiful interface. nuff said.