HN user
Thanks, sounds great! Not sure if you are part of the research team but a follow up question nevertheless. Learning from JpegXL, what would it take to develop another widely supported image format? Would the research stage already need to be carried out as a multi-corporate effort?
Sorry for perhaps missing it but it states "It provides both a fully interoperable encoder and decoder complying with the original JPEG standard". Does that mean that jpegli-encoded images can be decoded by all jpeg decoders? But it will not have the same quality?
Front page worthy effort?
Absolutely is not. There was nothing AI in this, simple classical image processing tools only.
Things like this are not binary "can do/cannot do". Sure some kind of estimate could be made. But it is challenging when the scene is not static, the speed of the camera varies, lack of texture, motion blur... stereo is not without its problems of course. Lidars seem to be good in lot of the places where visible light cameras are struggling.
I agree that when Tesla made their first promises of delivering FSD HW (just add SW a little bit later) reasonably priced solid state lidars were not available. So a story had to be made 'humans have eyes and they can drive, lidar is not needed'. Super hard to back now when they sold FSD to so many customers still waiting for the SW update.
Do you have personal experience with commaAI's pilot? I don't so basing all on videos alone. In a video to promote the bleeding edge of their development with Lex Fridman Hotz was driving the car to a specific part of a freeway that had a tilt and he was able to let go off the wheel for a brief moment. In another part the car was driving by itself in an urban setting and the highlight was that it was able to stop by itself at an intersection. This is light years removed from anything resembling a self driving system. But again, if you have more insight than the few videos I am basing this on please share. Would be interesting to hear more.
In this case it’s more than HU value LUT though. Photon counting allows per energy bin attenuation to capture more information. It’s cool and promising but costly and not new per se.
Pretty much. Normal x-ray sensors are like camera sensors and produce an estimate of total energy of the photons / pixel. Photon counting sensors are able to estimate how many photons of certain energy range hit the sensor. So instead of value ’1023’ you might get ’63’ of < 20kv, ’500’ of 20-50kv and ’50’ of > 50kv. So poor mans spectroscopy. Naive way to produce ’color images’ would be to put those buckets to red green and blue channel and now different materials along the beam path might yield equal values with normal x-ray sensors but slightly different colors with photon counting sensors. edit: clarified last sentence
Photon counting sensors have been available for quite a while. This is unfortunately not new. It’s not ’color’ it’s multi channel imaging. X-ray tubes produce wide spectrum of x-rays, these type of sensors are able to put different bands to different buckets.
Coloring is arbitrary and conveniently chosen in the example to look like visible wavelength response as if the patient was dissected. It does not work for more complicated anatomy and even in this case needs very likely post processing.
I wish we could have a peek inside the patient as if we had opened it up but this is not yet that.
Thank you so much for your reply. Appreciate it!
Hi! I just recently moved back from US where I worked with an O-1 VISA. Since I’ve already been once classified eligible for O-1 does that offer any shortcuts for getting another O-1 with potentially a different employer?