sparse fourier transforms
play

Sparse Fourier Transforms Eric Price UT Austin Eric Price Sparse - PowerPoint PPT Presentation

Jul 18, 2023 •129 likes •1.94k views

Sparse Fourier Transforms Eric Price UT Austin Eric Price Sparse Fourier Transforms 1 / 36 The Fourier Transform Conversion between time and frequency domains Frequency Domain Time Domain Fourier Transform Displacement of Air Concert A

  1. Generic Algorithm Outline 1-sparse recovery 1-sparse recovery x x ′ O ( k ) � Permute Filters 1-sparse recovery 1-sparse recovery Algorithm for k = 1 (exact or approximate) 1 Method to reduce to k = 1 case 2 ◮ Split � x into O ( k ) “random” parts Eric Price Sparse Fourier Transforms 8 / 36

  2. Generic Algorithm Outline 1-sparse recovery 1-sparse recovery x x ′ O ( k ) � Permute Filters 1-sparse recovery 1-sparse recovery Algorithm for k = 1 (exact or approximate) 1 Method to reduce to k = 1 case 2 ◮ Split � x into O ( k ) “random” parts ◮ Can sample time domain of the parts. Eric Price Sparse Fourier Transforms 8 / 36

  3. Generic Algorithm Outline 1-sparse recovery 1-sparse recovery x x ′ O ( k ) � Permute Filters 1-sparse recovery 1-sparse recovery Algorithm for k = 1 (exact or approximate) 1 Method to reduce to k = 1 case 2 ◮ Split � x into O ( k ) “random” parts ◮ Can sample time domain of the parts. ⋆ O ( k log k ) time to get one sample from each of the k parts. Eric Price Sparse Fourier Transforms 8 / 36

  4. Generic Algorithm Outline 1-sparse recovery 1-sparse recovery x x ′ O ( k ) � Permute Filters 1-sparse recovery 1-sparse recovery Algorithm for k = 1 (exact or approximate) 1 Method to reduce to k = 1 case 2 ◮ Split � x into O ( k ) “random” parts ◮ Can sample time domain of the parts. ⋆ O ( k log k ) time to get one sample from each of the k parts. Finds “most” of signal; repeat on residual 3 Eric Price Sparse Fourier Transforms 8 / 36

  5. Talk Outline Algorithm for k = 1 1 Eric Price Sparse Fourier Transforms 9 / 36

  6. Talk Outline Algorithm for k = 1 1 Reducing k to 1 2 Eric Price Sparse Fourier Transforms 9 / 36

  7. Talk Outline Algorithm for k = 1 1 Reducing k to 1 2 3 Putting it together Eric Price Sparse Fourier Transforms 9 / 36

  8. Talk Outline Algorithm for k = 1 1 Reducing k to 1 2 3 Putting it together Continuous setting 4 Eric Price Sparse Fourier Transforms 9 / 36

  9. Talk Outline Algorithm for k = 1 1 Reducing k to 1 2 3 Putting it together Continuous setting 4 Eric Price Sparse Fourier Transforms 10 / 36

  10. Algorithm for k = 1: one dimension, exact case Lemma a We can compute a 1 -sparse � x in O ( 1 ) time. � x : � a if i = t � x i = t 0 otherwise Eric Price Sparse Fourier Transforms 11 / 36

  11. Algorithm for k = 1: one dimension, exact case Lemma a We can compute a 1 -sparse � x in O ( 1 ) time. � x : � a if i = t � x i = t 0 otherwise Then x = ( a , a ω t , a ω 2 t , a ω 3 t , . . . , a ω ( n − 1 ) t ) . Eric Price Sparse Fourier Transforms 11 / 36

  12. Algorithm for k = 1: one dimension, exact case Lemma a We can compute a 1 -sparse � x in O ( 1 ) time. � x : � a if i = t � x i = t 0 otherwise Then x = ( a , a ω t , a ω 2 t , a ω 3 t , . . . , a ω ( n − 1 ) t ) . x 0 = a Eric Price Sparse Fourier Transforms 11 / 36

  13. Algorithm for k = 1: one dimension, exact case Lemma a We can compute a 1 -sparse � x in O ( 1 ) time. � x : � a if i = t � x i = t 0 otherwise Then x = ( a , a ω t , a ω 2 t , a ω 3 t , . . . , a ω ( n − 1 ) t ) . x 1 = a ω t x 0 = a Eric Price Sparse Fourier Transforms 11 / 36

  14. Algorithm for k = 1: one dimension, exact case Lemma a We can compute a 1 -sparse � x in O ( 1 ) time. � x : � a if i = t � x i = t 0 otherwise Then x = ( a , a ω t , a ω 2 t , a ω 3 t , . . . , a ω ( n − 1 ) t ) . x 1 = a ω t x 0 = a x 1 / x 0 = ω t = ⇒ t . Eric Price Sparse Fourier Transforms 11 / 36

  15. Algorithm for k = 1: one dimension, exact case Lemma a We can compute a 1 -sparse � x in O ( 1 ) time. � x : � a if i = t � x i = t 0 otherwise Then x = ( a , a ω t , a ω 2 t , a ω 3 t , . . . , a ω ( n − 1 ) t ) . x 1 = a ω t x 0 = a x 1 / x 0 = ω t = ⇒ t . � Eric Price Sparse Fourier Transforms 11 / 36

  16. Algorithm for k = 1: one dimension, exact case Lemma a We can compute a 1 -sparse � x in O ( 1 ) time. � x : � a if i = t � x i = t 0 otherwise Then x = ( a , a ω t , a ω 2 t , a ω 3 t , . . . , a ω ( n − 1 ) t ) . x 1 = a ω t x 0 = a x 1 / x 0 = ω t = ⇒ t . � (Related to OFDM, Prony’s method, matrix pencil.) Eric Price Sparse Fourier Transforms 11 / 36

  17. Algorithm for k = 1: one dimension, approximate case Lemma Suppose � x is approximately 1 -sparse: | � x t | / � � x � 2 � 90 %. Then we can recover it with O ( log n ) samples and O ( log 2 n ) time. Eric Price Sparse Fourier Transforms 12 / 36

  18. Algorithm for k = 1: one dimension, approximate case Lemma Suppose � x is approximately 1 -sparse: | � x t | / � � x � 2 � 90 %. Then we can recover it with O ( log n ) samples and O ( log 2 n ) time. x 1 / x 0 = ω t With exact sparsity: log n bits in a single measurement. Eric Price Sparse Fourier Transforms 12 / 36

  19. Algorithm for k = 1: one dimension, approximate case Lemma Suppose � x is approximately 1 -sparse: | � x t | / � � x � 2 � 90 %. Then we can recover it with O ( log n ) samples and O ( log 2 n ) time. x 1 / x 0 = ω t + noise With exact sparsity: log n bits in a single measurement. With noise: only constant number of useful bits. Eric Price Sparse Fourier Transforms 12 / 36

  20. Algorithm for k = 1: one dimension, approximate case Lemma Suppose � x is approximately 1 -sparse: | � x t | / � � x � 2 � 90 %. Then we can recover it with O ( log n ) samples and O ( log 2 n ) time. x 1 / x 0 = ω t + noise With exact sparsity: log n bits in a single measurement. With noise: only constant number of useful bits. Choose Θ ( log n ) time shifts c to recover i . Eric Price Sparse Fourier Transforms 12 / 36

  21. Algorithm for k = 1: one dimension, approximate case Lemma Suppose � x is approximately 1 -sparse: | � x t | / � � x � 2 � 90 %. Then we can recover it with O ( log n ) samples and O ( log 2 n ) time. x c 2 / x 0 = ω c 2 t + noise With exact sparsity: log n bits in a single measurement. With noise: only constant number of useful bits. Choose Θ ( log n ) time shifts c to recover i . Eric Price Sparse Fourier Transforms 12 / 36

  22. Algorithm for k = 1: one dimension, approximate case Lemma Suppose � x is approximately 1 -sparse: | � x t | / � � x � 2 � 90 %. Then we can recover it with O ( log n ) samples and O ( log 2 n ) time. x c 3 / x 0 = ω c 3 t + noise With exact sparsity: log n bits in a single measurement. With noise: only constant number of useful bits. Choose Θ ( log n ) time shifts c to recover i . Eric Price Sparse Fourier Transforms 12 / 36

  23. Algorithm for k = 1: one dimension, approximate case Lemma Suppose � x is approximately 1 -sparse: | � x t | / � � x � 2 � 90 %. Then we can recover it with O ( log n ) samples and O ( log 2 n ) time. x c 3 / x 0 = ω c 3 t + noise With exact sparsity: log n bits in a single measurement. With noise: only constant number of useful bits. Choose Θ ( log n ) time shifts c to recover i . Error correcting code with efficient recovery = ⇒ lemma. � Eric Price Sparse Fourier Transforms 12 / 36

  24. Talk Outline Algorithm for k = 1 1 Reducing k to 1 2 3 Putting it together Continuous setting 4 Eric Price Sparse Fourier Transforms 13 / 36

  25. Algorithm for general k Reduce general k to k = 1. 1-sparse recovery 1-sparse recovery x x ′ O ( k ) � Permute Filters 1-sparse recovery 1-sparse recovery Eric Price Sparse Fourier Transforms 14 / 36

  26. Algorithm for general k Reduce general k to k = 1. “Filters”: partition frequencies into O ( k ) buckets. 1-sparse recovery 1-sparse recovery x x ′ O ( k ) � Permute Filters 1-sparse recovery 1-sparse recovery Eric Price Sparse Fourier Transforms 14 / 36

  27. Algorithm for general k Reduce general k to k = 1. “Filters”: partition frequencies into O ( k ) buckets. 1-sparse recovery 1-sparse recovery x x ′ O ( k ) � Permute Filters 1-sparse recovery 1-sparse recovery Eric Price Sparse Fourier Transforms 14 / 36

  28. Algorithm for general k Reduce general k to k = 1. “Filters”: partition frequencies into O ( k ) buckets. 1-sparse recovery 1-sparse recovery x x ′ O ( k ) � Permute Filters 1-sparse recovery 1-sparse recovery Eric Price Sparse Fourier Transforms 14 / 36

  29. Algorithm for general k Reduce general k to k = 1. “Filters”: partition frequencies into O ( k ) buckets. ◮ Sample from time domain of each bucket with O ( log n ) overhead. 1-sparse recovery 1-sparse recovery x x ′ O ( k ) � Permute Filters 1-sparse recovery 1-sparse recovery Eric Price Sparse Fourier Transforms 14 / 36

  30. Algorithm for general k Reduce general k to k = 1. “Filters”: partition frequencies into O ( k ) buckets. ◮ Sample from time domain of each bucket with O ( log n ) overhead. ◮ Recovered by k = 1 algorithm 1-sparse recovery 1-sparse recovery x x ′ O ( k ) � Permute Filters 1-sparse recovery 1-sparse recovery Eric Price Sparse Fourier Transforms 14 / 36

  31. Algorithm for general k Reduce general k to k = 1. “Filters”: partition frequencies into O ( k ) buckets. ◮ Sample from time domain of each bucket with O ( log n ) overhead. ◮ Recovered by k = 1 algorithm Most frequencies alone in bucket. 1-sparse recovery 1-sparse recovery x x ′ O ( k ) � Permute Filters 1-sparse recovery 1-sparse recovery Eric Price Sparse Fourier Transforms 14 / 36

  32. Algorithm for general k Reduce general k to k = 1. “Filters”: partition frequencies into O ( k ) buckets. ◮ Sample from time domain of each bucket with O ( log n ) overhead. ◮ Recovered by k = 1 algorithm Most frequencies alone in bucket. 1-sparse recovery 1-sparse recovery x x ′ O ( k ) � Permute Filters 1-sparse recovery 1-sparse recovery Eric Price Sparse Fourier Transforms 14 / 36

  33. Algorithm for general k Reduce general k to k = 1. “Filters”: partition frequencies into O ( k ) buckets. ◮ Sample from time domain of each bucket with O ( log n ) overhead. ◮ Recovered by k = 1 algorithm Most frequencies alone in bucket. 1-sparse recovery 1-sparse recovery x x ′ O ( k ) � Permute Filters 1-sparse recovery 1-sparse recovery Eric Price Sparse Fourier Transforms 14 / 36

  34. Algorithm for general k Reduce general k to k = 1. “Filters”: partition frequencies into O ( k ) buckets. ◮ Sample from time domain of each bucket with O ( log n ) overhead. ◮ Recovered by k = 1 algorithm Most frequencies alone in bucket. 1-sparse recovery 1-sparse recovery x x ′ O ( k ) � Permute Filters 1-sparse recovery 1-sparse recovery Eric Price Sparse Fourier Transforms 14 / 36

  35. Algorithm for general k Reduce general k to k = 1. “Filters”: partition frequencies into O ( k ) buckets. ◮ Sample from time domain of each bucket with O ( log n ) overhead. ◮ Recovered by k = 1 algorithm Most frequencies alone in bucket. Random permutation 1-sparse recovery 1-sparse recovery x x ′ O ( k ) � Permute Filters 1-sparse recovery 1-sparse recovery Eric Price Sparse Fourier Transforms 14 / 36

  36. Algorithm for general k Reduce general k to k = 1. “Filters”: partition frequencies into O ( k ) buckets. ◮ Sample from time domain of each bucket with O ( log n ) overhead. ◮ Recovered by k = 1 algorithm Most frequencies alone in bucket. Random permutation 1-sparse recovery 1-sparse recovery x x ′ O ( k ) � Permute Filters 1-sparse recovery 1-sparse recovery Eric Price Sparse Fourier Transforms 14 / 36

  37. Algorithm for general k Reduce general k to k = 1. “Filters”: partition frequencies into O ( k ) buckets. ◮ Sample from time domain of each bucket with O ( log n ) overhead. ◮ Recovered by k = 1 algorithm Most frequencies alone in bucket. Random permutation 1-sparse recovery 1-sparse recovery x x ′ � Permute Filters O ( k ) 1-sparse recovery 1-sparse recovery Eric Price Sparse Fourier Transforms 14 / 36

  38. Algorithm for general k Reduce general k to k = 1. “Filters”: partition frequencies into O ( k ) buckets. ◮ Sample from time domain of each bucket with O ( log n ) overhead. ◮ Recovered by k = 1 algorithm Most frequencies alone in bucket. Random permutation 1-sparse recovery 1-sparse recovery x x ′ � Permute Filters O ( k ) 1-sparse recovery Recovers most of � x : 1-sparse recovery Lemma (Partial sparse recovery) x ′ such that In O ( k log n ) expected time, we can compute an estimate � x ′ is k / 2 -sparse. x − � � Eric Price Sparse Fourier Transforms 14 / 36

  39. Going from finding most coordinates to finding all � x Partial k -sparse recovery 1-sparse recovery 1-sparse recovery x x ′ � O ( k ) Permute Filters 1-sparse recovery 1-sparse recovery Lemma (Partial sparse recovery) x ′ such that In O ( k log n ) expected time, we can compute an estimate � x ′ is k / 2 -sparse. x − � � Eric Price Sparse Fourier Transforms 15 / 36

  40. Going from finding most coordinates to finding all x ′ x − � � Partial k -sparse recovery 1-sparse recovery 1-sparse recovery x x ′ � O ( k ) Permute Filters 1-sparse recovery 1-sparse recovery Lemma (Partial sparse recovery) x ′ such that In O ( k log n ) expected time, we can compute an estimate � x ′ is k / 2 -sparse. x − � � Eric Price Sparse Fourier Transforms 15 / 36

  41. Going from finding most coordinates to finding all x ′ x − � � Partial k -sparse recovery 1-sparse recovery 1-sparse recovery x x ′ � O ( k ) Permute Filters 1-sparse recovery 1-sparse recovery Lemma (Partial sparse recovery) x ′ such that In O ( k log n ) expected time, we can compute an estimate � x ′ is k / 2 -sparse. x − � � Repeat, k → k / 2 → k / 4 → · · · Eric Price Sparse Fourier Transforms 15 / 36

  42. Going from finding most coordinates to finding all x ′ x − � � Partial k -sparse recovery 1-sparse recovery 1-sparse recovery x x ′ � O ( k ) Permute Filters 1-sparse recovery 1-sparse recovery Lemma (Partial sparse recovery) x ′ such that In O ( k log n ) expected time, we can compute an estimate � x ′ is k / 2 -sparse. x − � � Repeat, k → k / 2 → k / 4 → · · · Eric Price Sparse Fourier Transforms 15 / 36

  43. Going from finding most coordinates to finding all x ′ x − � � Partial k -sparse recovery 1-sparse recovery 1-sparse recovery x x ′ � O ( k ) Permute Filters 1-sparse recovery 1-sparse recovery Lemma (Partial sparse recovery) x ′ such that In O ( k log n ) expected time, we can compute an estimate � x ′ is k / 2 -sparse. x − � � Repeat, k → k / 2 → k / 4 → · · · Eric Price Sparse Fourier Transforms 15 / 36

  44. Going from finding most coordinates to finding all x ′ x − � � Partial k -sparse recovery 1-sparse recovery 1-sparse recovery x x ′ � O ( k ) Permute Filters 1-sparse recovery 1-sparse recovery Lemma (Partial sparse recovery) x ′ such that In O ( k log n ) expected time, we can compute an estimate � x ′ is k / 2 -sparse. x − � � Repeat, k → k / 2 → k / 4 → · · · Eric Price Sparse Fourier Transforms 15 / 36

  45. Going from finding most coordinates to finding all x ′ x − � � Partial k -sparse recovery 1-sparse recovery 1-sparse recovery x � x ′ O ( k ) Permute Filters 1-sparse recovery 1-sparse recovery Lemma (Partial sparse recovery) x ′ such that In O ( k log n ) expected time, we can compute an estimate � x ′ is k / 2 -sparse. x − � � Repeat, k → k / 2 → k / 4 → · · · Theorem We can compute � x in O ( k log n ) expected time. Eric Price Sparse Fourier Transforms 15 / 36

  46. Going from finding most coordinates to finding all x ′ x − � � Partial k -sparse recovery 1-sparse recovery 1-sparse recovery x � x ′ O ( k ) Permute Filters 1-sparse recovery 1-sparse recovery Lemma (Partial sparse recovery) x ′ such that In O ( k log n ) expected time, we can compute an estimate � x ′ is k / 2 -sparse. x − � � Repeat, k → k / 2 → k / 4 → · · · Theorem We can compute � x in O ( k log n ) expected time. Eric Price Sparse Fourier Transforms 15 / 36

  47. Talk Outline Algorithm for k = 1 1 Reducing k to 1 2 3 Putting it together Continuous setting 4 Eric Price Sparse Fourier Transforms 16 / 36

  48. How can you hope for sublinear time? Time Frequency n -dimensional DFT: O ( n log n ) x → � x × ∗ = = Eric Price Sparse Fourier Transforms 17 / 36

  49. How can you hope for sublinear time? Time Frequency n -dimensional DFT: O ( n log n ) x → � x × ∗ = = Eric Price Sparse Fourier Transforms 17 / 36

  50. How can you hope for sublinear time? Time Frequency n -dimensional DFT: O ( n log n ) x → � x × ∗ n -dimensional DFT of first k terms: O ( n log n ) x · rect → � x ∗ sinc. = = Eric Price Sparse Fourier Transforms 17 / 36

  51. How can you hope for sublinear time? Time Frequency n -dimensional DFT: O ( n log n ) x → � x × ∗ n -dimensional DFT of first k terms: O ( n log n ) x · rect → � x ∗ sinc. = = Eric Price Sparse Fourier Transforms 17 / 36

  52. How can you hope for sublinear time? Time Frequency n -dimensional DFT: O ( n log n ) x → � x × ∗ n -dimensional DFT of first k terms: O ( n log n ) x · rect → � x ∗ sinc. = = k -dimensional DFT of first k terms: O ( B log B ) alias ( x · rect ) → subsample ( � x ∗ sinc ) . Eric Price Sparse Fourier Transforms 17 / 36

  53. How can you hope for sublinear time? Time Frequency n -dimensional DFT: O ( n log n ) x → � x × ∗ n -dimensional DFT of first k terms: O ( n log n ) x · rect → � x ∗ sinc. = = k -dimensional DFT of first k terms: O ( B log B ) alias ( x · rect ) → subsample ( � x ∗ sinc ) . Eric Price Sparse Fourier Transforms 17 / 36

  54. Use a better filter GMS05, HIKP12, IKP14, IK14 Filter (time): Gaussian · sinc Filter (frequency): Gaussian * rectangle Filter: sparse F for which � F is large on an interval . Eric Price Sparse Fourier Transforms 18 / 36

  55. Use a better filter GMS05, HIKP12, IKP14, IK14 Filter (time): Gaussian · sinc Filter (frequency): Gaussian * rectangle Filter: sparse F for which � F is large on an interval . We can permute the frequencies: x ′ ⇒ � x i = � i = x σ i = x σ − 1 i Eric Price Sparse Fourier Transforms 18 / 36

  56. Use a better filter GMS05, HIKP12, IKP14, IK14 Filter (time): Gaussian · sinc Filter (frequency): Gaussian * rectangle Filter: sparse F for which � F is large on an interval . We can permute the frequencies: x ′ ⇒ � x i = � i = x σ i = x σ − 1 i Allows us to convert worst case to random case. Eric Price Sparse Fourier Transforms 18 / 36

  57. Algorithm for exactly sparse signals Original signal x Goal ˆ x Eric Price Sparse Fourier Transforms 19 / 36

  58. Algorithm for exactly sparse signals Computed F · x Filtered signal c F ∗ c x Eric Price Sparse Fourier Transforms 19 / 36

  59. Algorithm for exactly sparse signals F · x aliased to k terms Filtered signal c F ∗ c x Eric Price Sparse Fourier Transforms 19 / 36

  60. Algorithm for exactly sparse signals F · x aliased to k terms Computed samples of c F ∗ c x Eric Price Sparse Fourier Transforms 19 / 36

  61. Algorithm for exactly sparse signals F · x aliased to k terms Computed samples of c F ∗ c x Eric Price Sparse Fourier Transforms 19 / 36

  62. Algorithm for exactly sparse signals F · x aliased to k terms Knowledge about ˆ x Eric Price Sparse Fourier Transforms 19 / 36

  63. Algorithm for exactly sparse signals F · x aliased to k terms Knowledge about ˆ x Eric Price Sparse Fourier Transforms 19 / 36

  64. Algorithm for exactly sparse signals F · x aliased to k terms Knowledge about ˆ x Lemma If t is isolated in its bucket and in the “super-pass” region, the value b we compute for its bucket satisfies b = � x t . Computing the b for all O ( k ) buckets takes O ( k log n ) time. Eric Price Sparse Fourier Transforms 19 / 36

  65. Algorithm Lemma For most t, the value b we compute for its bucket satisfies b = � x t . Computing the b for all O ( k ) buckets takes O ( k log n ) time. Eric Price Sparse Fourier Transforms 20 / 36

  66. Algorithm Lemma For most t, the value b we compute for its bucket satisfies b = � x t . Computing the b for all O ( k ) buckets takes O ( k log n ) time. Time-shift x by one and repeat: b ′ = � x t ω t . Divide to get b ′ / b = ω t Eric Price Sparse Fourier Transforms 20 / 36

  67. Algorithm Lemma For most t, the value b we compute for its bucket satisfies b = � x t . Computing the b for all O ( k ) buckets takes O ( k log n ) time. Time-shift x by one and repeat: b ′ = � x t ω t . Divide to get b ′ / b = ω t = ⇒ can compute t . Eric Price Sparse Fourier Transforms 20 / 36

  68. Algorithm Lemma For most t, the value b we compute for its bucket satisfies b = � x t . Computing the b for all O ( k ) buckets takes O ( k log n ) time. Time-shift x by one and repeat: b ′ = � x t ω t . Divide to get b ′ / b = ω t = ⇒ can compute t . ◮ Just like our 1-sparse recovery algorithm, x 1 / x 0 = ω t . Eric Price Sparse Fourier Transforms 20 / 36

  69. Algorithm Lemma For most t, the value b we compute for its bucket satisfies b = � x t . Computing the b for all O ( k ) buckets takes O ( k log n ) time. Time-shift x by one and repeat: b ′ = � x t ω t . Divide to get b ′ / b = ω t = ⇒ can compute t . ◮ Just like our 1-sparse recovery algorithm, x 1 / x 0 = ω t . x ′ such that � x ′ is k / 2-sparse. Gives partial sparse recovery: � x − � 1-sparse recovery 1-sparse recovery x ′ x � Permute Filters O ( k ) 1-sparse recovery 1-sparse recovery Eric Price Sparse Fourier Transforms 20 / 36

  70. Algorithm Lemma For most t, the value b we compute for its bucket satisfies b = � x t . Computing the b for all O ( k ) buckets takes O ( k log n ) time. Time-shift x by one and repeat: b ′ = � x t ω t . Divide to get b ′ / b = ω t = ⇒ can compute t . ◮ Just like our 1-sparse recovery algorithm, x 1 / x 0 = ω t . x ′ such that � x ′ is k / 2-sparse. Gives partial sparse recovery: � x − � 1-sparse recovery 1-sparse recovery x ′ x � Permute Filters O ( k ) 1-sparse recovery 1-sparse recovery Repeat k → k / 2 → k / 4 → · · · Eric Price Sparse Fourier Transforms 20 / 36

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend

Signals and Systems Chapter 4: The Continuous Time Fourier Transform Derivation of the CT Fourier

Signals and Systems Chapter 4: The Continuous Time Fourier Transform Derivation of the CT Fourier

Signals and Systems Chapter 4: The Continuous Time Fourier Transform Derivation of the CT Fourier Transform pair Examples of Fourier Transforms Topic three Fourier Transforms of Periodic Signals Properties of the CT Fourier

426 views • 31 slides
Chapter 14: Fourier Transforms and Boundary Value Problems in an Unbounded Region Department of

Chapter 14: Fourier Transforms and Boundary Value Problems in an Unbounded Region Department of

Fourier Transforms BVPs in an Unbounded Region Chapter 14: Fourier Transforms and Boundary Value Problems in an Unbounded Region Department of Electrical Engineering National Taiwan University ihwang@ntu.edu.tw December 25, 2013 1 / 27

459 views • 27 slides
Lecture 5: Statistical models of images Today Review Fourier transforms Image statistics

Lecture 5: Statistical models of images Today Review Fourier transforms Image statistics

Lecture 5: Statistical models of images Today Review Fourier transforms Image statistics Texture synthesis The Discrete Fourier transform The Discrete Fourier Transform (DFT) transforms a signal f[n] into F[u] as: Fourier

853 views • 71 slides
Parallel Fast Fourier Transforms Gavin J. Pringle Joahcim Hein Introduction The Fourier

Parallel Fast Fourier Transforms Gavin J. Pringle Joahcim Hein Introduction The Fourier

Parallel Fast Fourier Transforms Gavin J. Pringle Joahcim Hein Introduction The Fourier Transform What, who, why? Mathematics and and its inherent properties Discrete Fourier Transform Fast Fourier Transform, or FFT

542 views • 42 slides
Concepts and Algorithms of Scientific and Visual Computing Discrete Fourier Transforms

Concepts and Algorithms of Scientific and Visual Computing Discrete Fourier Transforms

Concepts and Algorithms of Scientific and Visual Computing Discrete Fourier Transforms CS448J, Autumn 2015, Stanford University Dominik L. Michels Discrete-time Fourier Transform (DTFT) As in the continuous case, for x 2 ` 2 ( Z ), the

154 views • 12 slides
Lecture 3.7: Fourier transforms Matthew Macauley Department of Mathematical Sciences Clemson

Lecture 3.7: Fourier transforms Matthew Macauley Department of Mathematical Sciences Clemson

Lecture 3.7: Fourier transforms Matthew Macauley Department of Mathematical Sciences Clemson University http://www.math.clemson.edu/~macaule/ Math 4340, Advanced Engineering Mathematics M. Macauley (Clemson) Lecture 3.7: Fourier transforms

270 views • 7 slides
Fourier transforms of measures on the Brownian graph Jonathan M. Fraser The University of

Fourier transforms of measures on the Brownian graph Jonathan M. Fraser The University of

Fourier transforms of measures on the Brownian graph Jonathan M. Fraser The University of Manchester Joint work with Tuomas Sahlsten (Bristol, UK) and Tuomas Orponen (Helsinki, Finland) ICERM 10th March 2016 Jonathan M. Fraser Fourier

716 views • 57 slides
New Constructions of RIP Matrices with Fast Multiplication and Fewer Rows aka, sparse recovery

New Constructions of RIP Matrices with Fast Multiplication and Fewer Rows aka, sparse recovery

New Constructions of RIP Matrices with Fast Multiplication and Fewer Rows aka, sparse recovery from Fourier -like measurements with applications to fast Johnson-Lindenstrauss transforms , etc. Jelani Nelson, Eric Price, and Mary Wootters February

394 views • 24 slides
Linear Systems lecture 7 Fourier transforms academic year : 16-17 lecture : 7 UNIVERSITY OF

Linear Systems lecture 7 Fourier transforms academic year : 16-17 lecture : 7 UNIVERSITY OF

Linear Systems lecture 7 Fourier transforms academic year : 16-17 lecture : 7 UNIVERSITY OF TWENTE build : November 23, 2016 slides : 31 Today UNIVERSITY OF TWENTE The Fourier transform The Dirac delta function Properties of

1.16k views • 96 slides
Laplace Transform Prof. Seungchul Lee Industrial AI Lab. Two Famous Linear Transforms Fourier

Laplace Transform Prof. Seungchul Lee Industrial AI Lab. Two Famous Linear Transforms Fourier

Laplace Transform Prof. Seungchul Lee Industrial AI Lab. Two Famous Linear Transforms Fourier transform Laplace transform Pierre-Simon Laplace: Jean-Baptiste Joseph Fourier: 23 March 1749 5 March 1827) was a French scholar whose work was

495 views • 31 slides
Sparse Recovery and Fourier Sampling Eric Price MIT Eric Price (MIT) Sparse Recovery and

Sparse Recovery and Fourier Sampling Eric Price MIT Eric Price (MIT) Sparse Recovery and

Sparse Recovery and Fourier Sampling Eric Price MIT Eric Price (MIT) Sparse Recovery and Fourier Sampling 1 / 37 The Fourier Transform Conversion between time and frequency domains Frequency Domain Time Domain Fourier Transform

2.15k views • 187 slides
Fourier Transforms and Spectral Analysis Chaiwoot Boonyasiriwat September 17, 2020

Fourier Transforms and Spectral Analysis Chaiwoot Boonyasiriwat September 17, 2020

Fourier Transforms and Spectral Analysis Chaiwoot Boonyasiriwat September 17, 2020 Discrete-Time Fourier Transform The discrete-time Fourier transform (DTFT) of a discrete-time signal x ( k ) with sampling interval T is the Z transform X ( z

1.03k views • 80 slides
6.003: Signals and Systems Applications of Fourier Transforms November 17, 2011 1 Filtering Notion

6.003: Signals and Systems Applications of Fourier Transforms November 17, 2011 1 Filtering Notion

6.003: Signals and Systems Applications of Fourier Transforms November 17, 2011 1 Filtering Notion of a filter. LTI systems cannot create new frequencies. can only scale magnitudes and shift phases of existing components. Example: LowPass

937 views • 44 slides
Sparse Time-Frequency Transforms and Applications. Bruno Torr esani

Sparse Time-Frequency Transforms and Applications. Bruno Torr esani

Sparse Time-Frequency Transforms and Applications. Bruno Torr esani http://www.cmi.univ-mrs.fr/~torresan LATP, Universit e de Provence, Marseille DAFx, Montreal, September 2006 B. Torr esani (LATP Marseille) Sparse Time-Frequency

1.19k views • 53 slides
Divide and Conquer: The transform named in his honor is a mathematical technique that can be

Divide and Conquer: The transform named in his honor is a mathematical technique that can be

3/7/2016 Fourier Transforms Joseph Fourier observed that any continuous function f(x) can be expressed as a sum of sine functions sin( x + ), each one suitably amplified and shifted in phase. CSE373: Data Structures and

297 views • 3 slides
Operator valued Fourier transforms on nilpotent Lie groups Daniel Beltit a Institute of

Operator valued Fourier transforms on nilpotent Lie groups Daniel Beltit a Institute of

Operator valued Fourier transforms on nilpotent Lie groups Daniel Beltit a Institute of Mathematics of the Romanian Academy *** Joint work with Ingrid Beltit a (IMAR) and Jean Ludwig (UL) Hamburg, 16.02.2015 Daniel Beltit

458 views • 14 slides
Fast Fourier Transform Dima Kochkov 1 1 Department of Physics University of Illinois at

Fast Fourier Transform Dima Kochkov 1 1 Department of Physics University of Illinois at

Impact Math part Algorithm References Fast Fourier Transform Dima Kochkov 1 1 Department of Physics University of Illinois at Urbana-Champaign Algorithm interest meeting, 2016 Impact Math part Algorithm References Outline Summation

517 views • 25 slides
Integer multiplication with generalized Fermat primes Svyatoslav Covanov CARAMEL Team, LORIA,

Integer multiplication with generalized Fermat primes Svyatoslav Covanov CARAMEL Team, LORIA,

Integer multiplication with generalized Fermat primes Svyatoslav Covanov CARAMEL Team, LORIA, University of Lorraine Supervised by: Emmanuel Thom and Jrmie Detrey Journes nationales du Calcul Formel 2015 (Cluny) November 4, 2015

670 views • 44 slides
Fast Fourier Transform 2 Announcements HW 3 posted tonight (after this) 3 Fast Fourier

Fast Fourier Transform 2 Announcements HW 3 posted tonight (after this) 3 Fast Fourier

1 Fast Fourier Transform 2 Announcements HW 3 posted tonight (after this) 3 Fast Fourier Transform 4 Math ground work Let Called n th roots of unity We will prove/use: 5 Math ground work Prove: By definition: Prove: Again, by

490 views • 30 slides
Complexity Theory of Polynomial-Time Problems Lecture 6: 3SUM Part I Karl Bringmann 3SUM given

Complexity Theory of Polynomial-Time Problems Lecture 6: 3SUM Part I Karl Bringmann 3SUM given

Complexity Theory of Polynomial-Time Problems Lecture 6: 3SUM Part I Karl Bringmann 3SUM given sets ,, of integers are there , , such that + + = 0 ? (we assume that we can

880 views • 47 slides
Fast Fourier transform (FFT) MATLAB tutorial series (Part 2.1) Pouyan Ebrahimbabaie Laboratory

Fast Fourier transform (FFT) MATLAB tutorial series (Part 2.1) Pouyan Ebrahimbabaie Laboratory

Fast Fourier transform (FFT) MATLAB tutorial series (Part 2.1) Pouyan Ebrahimbabaie Laboratory for Signal and Image Exploitation (INTELSIG) Dept. of Electrical Engineering and Computer Science University of Lige Lige, Belgium Applied

703 views • 51 slides
MATRICES AND LINEAR ALGEBRA Linear Algebra Matrix manipulation is the original essence of

MATRICES AND LINEAR ALGEBRA Linear Algebra Matrix manipulation is the original essence of

MATRICES AND LINEAR ALGEBRA Linear Algebra Matrix manipulation is the original essence of Matlab; hence the name MATrix LABoratory. In this section we will cover the basics of linear algebra, the ways of using Matlab in this context, and

1.08k views • 28 slides
learning and computer vision Sample slides only Presenter: Prof. Ioannis Pitas Aristotle

learning and computer vision Sample slides only Presenter: Prof. Ioannis Pitas Aristotle

Fast Convolution Algorithms for deep learning and computer vision Sample slides only Presenter: Prof. Ioannis Pitas Aristotle University of Thessaloniki pitas@csd.auth.gr Outline 1D convolutions Linear & Cyclic 1D convolutions

1.01k views • 24 slides
Shors Algorithm for Factoring: Background Quantum Algorithms In 1994, Peter Shor came up with O

Shors Algorithm for Factoring: Background Quantum Algorithms In 1994, Peter Shor came up with O

Shors Algorithm for Factoring: Background Quantum Algorithms In 1994, Peter Shor came up with O ( n 3 ) -time algorithm DPV Chapter 10 for factoring n -bit integers on a quantum algorithm. Why is this a big deal? O ( e n 1/3 ( log n )

712 views • 5 slides

More recommend