Patents by Inventor Jan Xu
Jan Xu has filed for patents to protect the following inventions. This listing includes patent applications that are pending as well as patents that have already been granted by the United States Patent and Trademark Office (USPTO).
-
Patent number: 12323593Abstract: A computer-implemented method for lossy image or video compression, transmission and decoding, the method including the steps of: (i) receiving an input image at a first computer system; (ii) encoding the input image using a first trained neural network, using the first computer system, to produce a latent representation; (iii) quantizing the latent representation using the first computer system to produce a quantized latent; (iv) entropy encoding the quantized latent into a bitstream, using the first computer system; (v) transmitting the bitstream to a second computer system; (vi) the second computer system entropy decoding the bitstream to produce the quantized latent; (vii) the second computer system using a second trained neural network to produce an output image from the quantized latent, wherein the output image is an approximation of the input image.Type: GrantFiled: August 4, 2023Date of Patent: June 3, 2025Assignee: DEEP RENDER LTD.Inventors: Chri Besenbruch, Ciro Cursio, Christopher Finlay, Vira Koshkina, Alexander Lytchier, Jan Xu, Arsalan Zafar
-
Publication number: 20250173910Abstract: A method for lossy video encoding, transmission and decoding, the method comprising the steps of: receiving an input video at a first computer system; encoding an input frame of the input video to produce a latent representation; producing a quantized latent; producing a hyper-latent representation; producing a quantized hyper-latent; entropy encoding the quantized latent; transmitting the entropy encoded quantized latent and the quantized hyper-latent to a second computer system; decoding the quantized hyper-latent to produce a set of context variables, wherein the set of context variables comprise a temporal context variable; entropy decoding the entropy encoded quantized latent using the set of context variables to obtain an output quantized latent; and decoding the output quantized latent to produce an output frame, wherein the output frame is an approximation of the input frame.Type: ApplicationFiled: April 25, 2023Publication date: May 29, 2025Inventors: Chris FINLAY, Christian BESENBRUCH, Jan XU, Bilal ABBASI, Christian ETMANN, Arsalan ZAFAR, Sebastjan CIZEL, Vira KOSHKINA
-
Patent number: 12256075Abstract: There is disclosed a computer-implemented method for lossy image or video compression, transmission and decoding, the method including the steps of: (i) receiving an input image at a first computer system; (ii) encoding the input image using a first trained neural network, using the first computer system, to produce a latent representation; (iii) quantizing the latent representation using the first computer system to produce a quantized latent; (iv) entropy encoding the quantized latent into a bitstream, using the first computer system; (v) transmitting the bitstream to a second computer system; (vi) the second computer system entropy decoding the bitstream to produce the quantized latent; (vii) the second computer system using a second trained neural network to produce an output image from the quantized latent, wherein the output image is an approximation of the input image. Related computer-implemented methods, systems, computer-implemented training methods and computer program products are disclosed.Type: GrantFiled: November 15, 2022Date of Patent: March 18, 2025Assignee: DEEP RENDER LTD.Inventors: Chri Besenbruch, Ciro Cursio, Christopher Finlay, Vira Koshkina, Alexander Lytchier, Jan Xu, Arsalan Zafar
-
Publication number: 20250086843Abstract: A method for lossy image and video encoding, transmission and decoding, the method comprising the steps of: receiving an input image at a first computer system; encoding the input image using a first trained neural network to produce a latent representation; performing a quantization process on the latent representation to produce a quantized latent; transmitting the quantized latent to a second computer system; decoding the quantized latent using a denoising process to produce an output image, wherein the output image is an approximation of the input image.Type: ApplicationFiled: December 21, 2022Publication date: March 13, 2025Inventors: Arsalan ZAFAR, Jan XU, Christiain BESENBRUCH, Bilal ABBASI, Aleksandar CHERGANSKI, Chris FINLAY, Christian ETMANN
-
Patent number: 12160579Abstract: There is disclosed a computer-implemented method for lossy image or video compression, transmission and decoding, the method including the steps of: (i) receiving an input image at a first computer system; (ii) encoding the input image using a first trained neural network, using the first computer system, to produce a latent representation; (iii) quantizing the latent representation using the first computer system to produce a quantized latent; (iv) entropy encoding the quantized latent into a bitstream, using the first computer system; (v) transmitting the bitstream to a second computer system; (vi) the second computer system entropy decoding the bitstream to produce the quantized latent; (vii) the second computer system using a second trained neural network to produce an output image from the quantized latent, wherein the output image is an approximation of the input image. Related computer-implemented methods, systems, computer-implemented training methods and computer program products are disclosed.Type: GrantFiled: August 4, 2023Date of Patent: December 3, 2024Assignee: DEEP RENDER LTD.Inventors: Chri Besenbruch, Ciro Cursio, Christopher Finlay, Vira Koshkina, Alexander Lytchier, Jan Xu, Arsalan Zafar
-
Publication number: 20240354553Abstract: A method for lossy image and video encoding, transmission and decoding, the method comprising the steps of: receiving an input image at a first computer system; encoding the input image using a first trained neural network to produce a latent representation; performing a quantization process on the latent representation to produce a quantized latent, wherein the sizes of the bins used in the quantization process are based on the input image; transmitting the quantized latent to a second computer system; decoding the quantized latent using a second trained neural network to produce an output image, wherein the output image is an approximation of the input image.Type: ApplicationFiled: August 3, 2022Publication date: October 24, 2024Inventors: Chris FINLAY, Jan XU, Christiain BESENBRUCH, Arsalan ZAFAR
-
Publication number: 20240312069Abstract: A computer-implemented method of training an image generative network f? for a set of training images, in which an output image {circumflex over (x)} is generated from an input image x of the set of training images non-losslessly, and in which a proxy network is trained for a gradient intractable perceptual metric that evaluates a quality of an output image {circumflex over (x)} given an input image x, the method of training using a plurality of scales for input images from the set of training images. In an embodiment, a blindspot network b? is trained which generates an output image {tilde over (x)} from an input image x. Related computer systems, computer program products and computer-implemented methods of training are disclosed.Type: ApplicationFiled: January 10, 2024Publication date: September 19, 2024Inventors: Chri BESENBRUCH, Ciro CURSIO, Christopher FINLAY, Vira KOSHKINA, Alexander LYTCHIER, Jan XU, Arsalan ZAFAR
-
Patent number: 12095994Abstract: There is disclosed a computer-implemented method for lossy image or video compression, transmission and decoding, the method including the steps of: (i) receiving an input image at a first computer system; (ii) encoding the input image using a first trained neural network, using the first computer system, to produce a latent representation; (iii) quantizing the latent representation using the first computer system to produce a quantized latent; (iv) entropy encoding the quantized latent into a bitstream, using the first computer system; (v) transmitting the bitstream to a second computer system; (vi) the second computer system entropy decoding the bitstream to produce the quantized latent; (vii) the second computer system using a second trained neural network to produce an output image from the quantized latent, wherein the output image is an approximation of the input image. Related computer-implemented methods, systems, computer-implemented training methods and computer program products are disclosed.Type: GrantFiled: August 4, 2023Date of Patent: September 17, 2024Assignee: DEEP RENDER LTD.Inventors: Chri Besenbruch, Ciro Cursio, Christopher Finlay, Vira Koshkina, Alexander Lytchier, Jan Xu, Arsalan Zafar
-
Patent number: 12081759Abstract: There is disclosed a computer-implemented method for lossy image or video compression, transmission and decoding, the method including the steps of: (i) receiving an input image at a first computer system; (ii) encoding the input image using a first trained neural network, using the first computer system, to produce a latent representation; (iii) quantizing the latent representation using the first computer system to produce a quantized latent; (iv) entropy encoding the quantized latent into a bitstream, using the first computer system; (v) transmitting the bitstream to a second computer system; (vi) the second computer system entropy decoding the bitstream to produce the quantized latent; (vii) the second computer system using a second trained neural network to produce an output image from the quantized latent, wherein the output image is an approximation of the input image. Related computer-implemented methods, systems, computer-implemented training methods and computer program products are disclosed.Type: GrantFiled: August 4, 2023Date of Patent: September 3, 2024Assignee: DEEP RENDER LTD.Inventors: Chri Besenbruch, Ciro Cursio, Christopher Finlay, Vira Koshkina, Alexander Lytchier, Jan Xu, Arsalan Zafar
-
Patent number: 12075053Abstract: There is disclosed a computer-implemented method for lossy image or video compression, transmission and decoding, the method including the steps of: (i) receiving an input image at a first computer system; (ii) encoding the input image using a first trained neural network, using the first computer system, to produce a latent representation; (iii) quantizing the latent representation using the first computer system to produce a quantized latent; (iv) entropy encoding the quantized latent into a bitstream, using the first computer system; (v) transmitting the bitstream to a second computer system; (vi) the second computer system entropy decoding the bitstream to produce the quantized latent; (vii) the second computer system using a second trained neural network to produce an output image from the quantized latent, wherein the output image is an approximation of the input image. Related computer-implemented methods, systems, computer-implemented training methods and computer program products are disclosed.Type: GrantFiled: August 4, 2023Date of Patent: August 27, 2024Assignee: DEEP RENDER LTD.Inventors: Chri Besenbruch, Ciro Cursio, Christopher Finlay, Vira Koshkina, Alexander Lytchier, Jan Xu, Arsalan Zafar
-
Patent number: 12028525Abstract: There is disclosed a computer-implemented method for lossy image or video compression, transmission and decoding, the method including the steps of: (i) receiving an input image at a first computer system; (ii) encoding the input image using a first trained neural network, using the first computer system, to produce a latent representation; (iii) quantizing the latent representation using the first computer system to produce a quantized latent; (iv) entropy encoding the quantized latent into a bitstream, using the first computer system; (v) transmitting the bitstream to a second computer system; (vi) the second computer system entropy decoding the bitstream to produce the quantized latent; (vii) the second computer system using a second trained neural network to produce an output image from the quantized latent, wherein the output image is an approximation of the input image. Related computer-implemented methods, systems, computer-implemented training methods and computer program products are disclosed.Type: GrantFiled: August 4, 2023Date of Patent: July 2, 2024Assignee: DEEP RENDER LTD.Inventors: Chri Besenbruch, Ciro Cursio, Christopher Finlay, Vira Koshkina, Alexander Lytchier, Jan Xu, Arsalan Zafar
-
Patent number: 12022077Abstract: There is disclosed a computer-implemented method for lossy image or video compression, transmission and decoding, the method including the steps of: (i) receiving an input image at a first computer system; (ii) encoding the input image using a first trained neural network, using the first computer system, to produce a latent representation; (iii) quantizing the latent representation using the first computer system to produce a quantized latent; (iv) entropy encoding the quantized latent into a bitstream, using the first computer system; (v) transmitting the bitstream to a second computer system; (vi) the second computer system entropy decoding the bitstream to produce the quantized latent; (vii) the second computer system using a second trained neural network to produce an output image from the quantized latent, wherein the output image is an approximation of the input image. Related computer-implemented methods, systems, computer-implemented training methods and computer program products are disclosed.Type: GrantFiled: August 4, 2023Date of Patent: June 25, 2024Assignee: DEEP RENDER LTD.Inventors: Chri Besenbruch, Ciro Cursio, Christopher Finlay, Vira Koshkina, Alexander Lytchier, Jan Xu, Arsalan Zafar
-
Patent number: 12015776Abstract: A computer-implemented method for lossy image or video compression, transmission and decoding, the method including the steps of: (i) receiving an input image at a first computer system; (ii) encoding the input image using a first trained neural network, using the first computer system, to produce a latent representation; (iii) quantizing the latent representation using the first computer system to produce a quantized latent; (iv) entropy encoding the quantized latent into a bitstream, using the first computer system; (v) transmitting the bitstream to a second computer system; (vi) the second computer system entropy decoding the bitstream to produce the quantized latent; (vii) the second computer system using a second trained neural network to produce an output image from the quantized latent, wherein the output image is an approximation of the input image. Related computer-implemented methods, systems, computer-implemented training methods and computer program products.Type: GrantFiled: August 4, 2023Date of Patent: June 18, 2024Assignee: DEEP RENDER LTD.Inventors: Chri Besenbruch, Ciro Cursio, Christopher Finlay, Vira Koshkina, Alexander Lytchier, Jan Xu, Arsalan Zafar
-
Publication number: 20240195971Abstract: There is disclosed a computer-implemented method for lossy image or video compression, transmission and decoding, the method including the steps of: (i) receiving an input image at a first computer system; (ii) encoding the input image using a first trained neural network, using the first computer system, to produce a latent representation; (iii) quantizing the latent representation using the first computer system to produce a quantized latent; (iv) entropy encoding the quantized latent into a bitstream, using the first computer system; (v) transmitting the bitstream to a second computer system; (vi) the second computer system entropy decoding the bitstream to produce the quantized latent; (vii) the second computer system using a second trained neural network to produce an output image from the quantized latent, wherein the output image is an approximation of the input image. Related computer-implemented methods, systems, computer-implemented training methods and computer program products are disclosed.Type: ApplicationFiled: August 4, 2023Publication date: June 13, 2024Inventors: Chri BESENBRUCH, Ciro CURSIO, Christopher FINLAY, Vira KOSHKINA, Alexander LYTCHIER, Jan XU, Arsalan ZAFAR
-
Patent number: 11985319Abstract: There is disclosed a computer-implemented method for lossy image or video compression, transmission and decoding, the method including the steps of: (i) receiving an input image at a first computer system; (ii) encoding the input image using a first trained neural network, using the first computer system, to produce a latent representation; (iii) quantizing the latent representation using the first computer system to produce a quantized latent; (iv) entropy encoding the quantized latent into a bitstream, using the first computer system; (v) transmitting the bitstream to a second computer system; (vi) the second computer system entropy decoding the bitstream to produce the quantized latent; (vii) the second computer system using a second trained neural network to produce an output image from the quantized latent, wherein the output image is an approximation of the input image. Related computer-implemented methods, systems, computer-implemented training methods and computer program products are disclosed.Type: GrantFiled: August 4, 2023Date of Patent: May 14, 2024Assignee: DEEP RENDER LTD.Inventors: Chri Besenbruch, Ciro Cursio, Christopher Finlay, Vira Koshkina, Alexander Lytchier, Jan Xu, Arsalan Zafar
-
Publication number: 20240107022Abstract: Lossy or lossless compression and transmission, comprising the steps of: (i) receiving an input image; (ii) encoding it to produce a y latent representation; (iii) encoding the y latent representation to produce a z hyperlatent representation; (iv) quantizing the z hyperlatent representation to produce a quantized z hyperlatent representation; (v) entropy encoding the quantized z hyperlatent representation into a first bitstream, (vi) processing the quantized z hyperlatent representation to obtain a location entropy parameter ?y, an entropy scale parameter ?y, and a context matrix Ay of the y latent representation; (vii) processing the y latent representation, the location entropy parameter py and the context matrix Ay, to obtain quantized latent residuals; (viii) entropy encoding the quantized latent residuals into a second bitstream; and (ix) transmitting the bitstreams.Type: ApplicationFiled: November 19, 2023Publication date: March 28, 2024Inventors: Chri BESENBRUCH, Aleksandar CHERGANSKI, Christopher FINLAY, Alexander LYTCHIER, Jonathan RAYNER, Tom RYDER, Jan XU, Arsalan ZAFAR
-
Patent number: 11936866Abstract: A method for lossy video encoding, transmission and decoding, the method comprising the steps of: receiving an input video at a first computer system; encoding an input frame of the input video to produce a latent representation; producing a quantized latent; producing a hyper-latent representation; producing a quantized hyper-latent; entropy encoding the quantized latent; transmitting the entropy encoded quantized latent and the quantized hyper-latent to a second computer system; decoding the quantized hyper-latent to produce a set of context variables, wherein the set of context variables comprise a temporal context variable; entropy decoding the entropy encoded quantized latent using the set of context variables to obtain an output quantized latent; and decoding the output quantized latent to produce an output frame, wherein the output frame is an approximation of the input frame.Type: GrantFiled: August 30, 2023Date of Patent: March 19, 2024Assignee: DEEP RENDER LTD.Inventors: Chris Finlay, Christian Besenbruch, Jan Xu, Bilal Abbasi, Christian Etmann, Arsalan Zafar, Sebastjan Cizel, Vira Koshkina
-
Publication number: 20240070925Abstract: A method of training one or more neural networks, the one or more neural networks being for use in lossy image or video encoding, transmission and decoding, the method comprising steps including: receiving an input image at a first computer system; encoding the input image using a first neural network and decoding the latent representation using a second neural network to produce an output image; at least one of the plurality of layers of the first or second neural network comprises a transformation; and the method further comprises the steps of: evaluating a difference between the output image and the input image and evaluating a function based on an output of the transformation; updating the parameters of the first neural network and the second neural network based on the evaluated difference and the evaluated function; and repeating the above steps.Type: ApplicationFiled: August 30, 2023Publication date: February 29, 2024Inventors: Chris FINLAY, Jonathan RAYNER, Jan XU, Christian BESENBRUCH, Arsalan ZAFAR, Sebastjan CIZEL, Vira KOSHKINA
-
Publication number: 20240056576Abstract: There is disclosed a computer-implemented method for lossy image or video compression, transmission and decoding, the method including the steps of: (i) receiving an input image at a first computer system; (ii) encoding the input image using a first trained neural network, using the first computer system, to produce a latent representation; (iii) quantizing the latent representation using the first computer system to produce a quantized latent; (iv) entropy encoding the quantized latent into a bitstream, using the first computer system; (v) transmitting the bitstream to a second computer system; (vi) the second computer system entropy decoding the bitstream to produce the quantized latent; (vii) the second computer system using a second trained neural network to produce an output image from the quantized latent, wherein the output image is an approximation of the input image. Related computer-implemented methods, systems, computer-implemented training methods and computer program products are disclosed.Type: ApplicationFiled: August 4, 2023Publication date: February 15, 2024Inventors: Chri BESENBRUCH, Ciro CURSIO, Christopher FINLAY, Vira KOSHKINA, Alexander LYTCHIER, Jan XU, Arsalan ZAFAR
-
Patent number: 11881003Abstract: A computer-implemented method of training an image generative network f? for a set of training images, in which an output image {circumflex over (x)} is generated from an input image x of the set of training images non-losslessly, and in which a proxy network is trained for a gradient intractable perceptual metric that evaluates a quality of an output image {circumflex over (x)} given an input image x, the method of training using a plurality of scales for input images from the set of training images. In an embodiment, a blindspot network b? is trained which generates an output image {tilde over (x)} from an input image x. Related computer systems, computer program products and computer-implemented methods of training are disclosed.Type: GrantFiled: January 20, 2023Date of Patent: January 23, 2024Assignee: DEEP RENDER LTD.Inventors: Chri Besenbruch, Ciro Cursio, Christopher Finlay, Vira Koshkina, Alexander Lytchier, Jan Xu, Arsalan Zafar