Home
last modified time | relevance | path

Searched full:cosine (Results 1 – 25 of 760) sorted by relevance

12345678910>>...31

/aosp_15_r20/external/aac/libFDK/include/
H A DFDK_trigFcts.h157 FIXP_DBL *cosine) { in fixp_sin_cos_residual_inline() argument
175 /* Cosine sign symmetry */ in fixp_sin_cos_residual_inline()
193 /* Cosine/Sine simetry for angles greater than PI/4 */ in fixp_sin_cos_residual_inline()
207 *cosine = (FIXP_DBL)((cl * csign) << (DFRACT_BITS - FRACT_BITS)); in fixp_sin_cos_residual_inline()
212 *cosine = (FIXP_DBL)(cl * csign) >> 1; in fixp_sin_cos_residual_inline()
220 * \brief Calculate cosine and sine value each of 2 angles different angle
230 FIXP_DBL residual, error0, error1, sine, cosine; in inline_fixp_cos_sin() local
231 residual = fixp_sin_cos_residual_inline(x1, scale, &sine, &cosine); in inline_fixp_cos_sin()
233 error1 = fMultDiv2(cosine, residual); in inline_fixp_cos_sin()
236 *out++ = cosine - (error0 << 1); in inline_fixp_cos_sin()
[all …]
/aosp_15_r20/external/tensorflow/tensorflow/python/keras/optimizer_v2/
H A Dlearning_rate_schedule.py556 """A LearningRateSchedule that uses a cosine decay schedule.
562 the training progresses. This schedule applies a cosine decay function
604 """Applies cosine decay to the learning rate.
650 """A LearningRateSchedule that uses a cosine decay schedule with restarts.
656 the training progresses. This schedule applies a cosine decay function with
698 """Applies cosine decay with restarts to the learning rate.
776 """A LearningRateSchedule that uses a linear cosine decay schedule.
785 Note that linear cosine decay is more aggressive than cosine decay and
789 the training progresses. This schedule applies a linear cosine decay
836 """Applies linear cosine decay to the learning rate.
[all …]
H A Dlegacy_learning_rate_decay.py456 """Applies cosine decay to the learning rate.
459 the training progresses. This function applies a cosine decay function
525 """Applies cosine decay with restarts to the learning rate.
528 the training progresses. This function applies a cosine decay function with
602 """Applies linear cosine decay to the learning rate.
604 Note that linear cosine decay is more aggressive than cosine decay and
608 the training progresses. This function applies a linear cosine decay function
636 num_periods: Number of periods in the cosine part of the decay. See
689 """Applies noisy linear cosine decay to the learning rate.
691 Note that linear cosine decay is more aggressive than cosine decay and
[all …]
/aosp_15_r20/frameworks/rs/script_api/
H A Drs_math.spec187 summary: Inverse cosine
189 Returns the inverse cosine, in radians.
210 summary: Inverse hyperbolic cosine
212 Returns the inverse hyperbolic cosine, in radians.
233 summary: Inverse cosine divided by pi
235 Returns the inverse cosine in radians, divided by pi.
237 To get an inverse cosine measured in degrees, use <code>acospi(a) * 180.f</code>.
616 summary: Cosine
618 Returns the cosine of an angle measured in radians.
639 summary: Hypebolic cosine
[all …]
/aosp_15_r20/external/aac/libFDK/src/
H A DFDK_trigFcts.cpp300 FIXP_DBL residual, error, sine, cosine; in fixp_cos() local
302 residual = fixp_sin_cos_residual_inline(x, scale, &sine, &cosine); in fixp_cos()
306 return cosine - error; in fixp_cos()
309 return SATURATE_LEFT_SHIFT(cosine - error, 1, DFRACT_BITS); in fixp_cos()
314 FIXP_DBL residual, error, sine, cosine; in fixp_sin() local
316 residual = fixp_sin_cos_residual_inline(x, scale, &sine, &cosine); in fixp_sin()
317 error = fMult(cosine, residual); in fixp_sin()
327 FIXP_DBL residual, error0, error1, sine, cosine; in fixp_cos_sin() local
329 residual = fixp_sin_cos_residual_inline(x, scale, &sine, &cosine); in fixp_cos_sin()
331 error1 = fMult(cosine, residual); in fixp_cos_sin()
[all …]
/aosp_15_r20/external/pytorch/torch/signal/windows/
H A Dwindows.py14 'cosine',
168 Computes a window with a simple cosine waveform, following the same implementation as SciPy.
171 The cosine window is defined as follows:
176 This formula differs from the typical cosine window formula by incorporating a 0.5 term in the nume…
196 >>> # Generates a symmetric cosine window.
197 >>> torch.signal.windows.cosine(10)
200 >>> # Generates a periodic cosine window.
201 >>> torch.signal.windows.cosine(10, sym=False)
207 def cosine( function
219 _window_function_checks('cosine', M, dtype, layout)
[all …]
/aosp_15_r20/frameworks/av/media/libaudioprocessing/
H A DAudioResamplerFirGen.h27 * As most of our functions use sine or cosine at equal steps,
32 * to calculate sine(wstart + n * wstep) or cosine(wstart + n * wstep)
37 * or looking at just the imaginary sine term, as the cosine follows identically:
44 * cosine generation due to the complex * complex multiply (full rotation).
52 SineGen(double wstart, double wstep, bool cosine = false) {
53 if (cosine) {
83 double mCurrent; // current value of sine/cosine
84 double mPrevious; // previous value of sine/cosine
101 SineGenGen(double outerStart, double outerStep, double innerStep, bool cosine = false)
102 : mSineInnerCur(outerStart, outerStep, cosine), in mSineInnerCur() argument
[all …]
/aosp_15_r20/external/pytorch/benchmarks/dynamo/
H A Dtorchbench.py410 cosine = self.args.cosine
414 return 1e-2, cosine
416 return 8 * 1e-2, cosine
417 return 1e-3, cosine
421 return 1e-2, cosine
425 if name in self._tolerance["cosine"]:
426 cosine = True
431 return tolerance, cosine
H A Dhuggingface.py492 cosine = self.args.cosine
500 return 2e-2, cosine
502 return 1e-2, cosine
505 return 4e-3, cosine
510 return 4e-3, cosine
511 return 1e-3, cosine
/aosp_15_r20/frameworks/base/location/java/android/location/
H A DKeplerianOrbitModel.java314 /** Amplitude of cosine harmonic correction term to angle of inclination in radians. */
320 /** Amplitude of cosine harmonic correction term to the orbit in meters. */
326 /** Amplitude of cosine harmonic correction term to the argument of latitude in radians. */
348 * Get the amplitude of cosine harmonic correction term to angle of inclination in radians.
363 /** Get the amplitude of cosine harmonic correction term to the orbit in meters. */
376 * Get the amplitude of cosine harmonic correction term to the argument of latitude in
455 * Sets the amplitude of cosine harmonic correction term to angle of inclination in
474 /** Sets the amplitude of cosine harmonic correction term to the orbit in meters. */
489 * Sets the amplitude of cosine harmonic correction term to the argument of latitude in
/aosp_15_r20/external/googleapis/google/cloud/aiplatform/v1beta1/
H A Dfeature_view.proto82 // Cosine Distance. Defined as 1 - cosine similarity.
85 // of COSINE distance. Our algorithms have been more optimized for
87 // mathematically equivalent to COSINE distance and results in the same
161 // Cosine Distance. Defined as 1 - cosine similarity.
164 // of COSINE distance. Our algorithms have been more optimized for
166 // mathematically equivalent to COSINE distance and results in the same
/aosp_15_r20/external/tensorflow/tensorflow/lite/g3doc/inference_with_metadata/task_library/
H A Dimage_embedder.md23 [cosine similarity](https://en.wikipedia.org/wiki/Cosine_similarity) between
57 // Compute cosine similarity.
92 # Compute cosine similarity.
105 Cosine similarity between normalized feature vectors return a score between -1
106 and 1. Higher is better, i.e. a cosine similarity of 1 means the two vectors are
110 Cosine similarity: 0.954312
H A Dtext_embedder.md24 [cosine similarity](https://en.wikipedia.org/wiki/Cosine_similarity) between
50 // Compute cosine similarity.
83 # Compute cosine similarity.
96 Cosine similarity between normalized feature vectors return a score between -1
97 and 1. Higher is better, i.e. a cosine similarity of 1 means the two vectors are
101 Cosine similarity: 0.954312
/aosp_15_r20/external/deqp-deps/glslang/Test/baseResults/
Dhlsl.intrinsics.vert.out17 0:5 arc cosine ( temp float)
41 0:16 cosine ( temp float)
43 0:17 hyp. cosine ( temp float)
122 0:49 cosine ( temp float)
166 0:72 arc cosine ( temp 2-component vector of float)
190 0:83 cosine ( temp 2-component vector of float)
192 0:84 hyp. cosine ( temp 2-component vector of float)
295 0:123 cosine ( temp 2-component vector of float)
331 0:141 arc cosine ( temp 3-component vector of float)
355 0:152 cosine ( temp 3-component vector of float)
[all …]
/aosp_15_r20/external/angle/third_party/glslang/src/Test/baseResults/
H A Dhlsl.intrinsics.vert.out17 0:5 arc cosine ( temp float)
41 0:16 cosine ( temp float)
43 0:17 hyp. cosine ( temp float)
122 0:49 cosine ( temp float)
166 0:72 arc cosine ( temp 2-component vector of float)
190 0:83 cosine ( temp 2-component vector of float)
192 0:84 hyp. cosine ( temp 2-component vector of float)
295 0:123 cosine ( temp 2-component vector of float)
331 0:141 arc cosine ( temp 3-component vector of float)
355 0:152 cosine ( temp 3-component vector of float)
[all …]
/aosp_15_r20/external/arm-optimized-routines/math/
H A Dsincosf.h17 /* The constants and polynomials for sine and cosine. */
23 double c0, c1, c2, c3, c4; /* Cosine polynomial. */
27 /* Polynomial data (the cosine polynomial is negated in the 2nd entry). */
40 /* Compute the sine and cosine of inputs X and X2 (X squared), using the
42 if odd the cosine and sine polynomials are swapped. */
71 N is the quadrant, and if odd the cosine polynomial is used. */
/aosp_15_r20/external/libxaac/decoder/armv7/
H A Dixheaacd_eld_decoder_sbr_pre_twiddle.s16 …LDR r6, [r2], #4 @Load and increment pointer *pTwiddles++ Lower - cosine , high…
19 SMULWB r8, r4, r6 @mult32x16in32(Xre, cosine)
24 …ADD r12, r8, r10, LSL #1 @mac32x16in32_shl( mult32x16in32_shl(Xre, cosine) , mult32x16…
30 SMULWB r9, r5, r6 @mult32x16in32(Xim, cosine)
37 …SUB r14, r9, r7, LSL #1 @sub32(mult32x16in32_shl(Xim, cosine) , mult32x16in32_shl(Xre,…
/aosp_15_r20/external/tensorflow/tensorflow/python/kernel_tests/signal/
H A Dwindow_ops_test.py41 """A simple implementation of a raised cosine window that matches SciPy.
49 a: The alpha parameter of the raised cosine window.
50 b: The beta parameter of the raised cosine window.
53 A raised cosine window of length `length`.
113 # The Hann window is a raised cosine window with parameters alpha=0.5 and
127 # The Hamming window is a raised cosine window with parameters alpha=0.54
/aosp_15_r20/packages/modules/AppSearch/framework/java/external/android/app/appsearch/ast/query/
DSemanticSearchNode.java90 * between embedding vectors. If no default is set, cosine similarity will be used.
109 * between embedding vectors. If no default is set, cosine similarity will be used.
130 * between embedding vectors. If no default is set, cosine similarity will be used.
219 * be {@code "COSINE"}.
224 * `semanticSearch(getEmbeddingParameter(0), -1.5, 2, "COSINE")`
287 return "\"COSINE\""; in getEmbeddingMetricString()
/aosp_15_r20/external/tensorflow/tensorflow/python/ops/
H A Dclustering_ops.py40 # Cosine distance between vectors U and V is defined as
42 COSINE_DISTANCE = 'cosine'
110 "squared_euclidean", "cosine".
163 Currently only Euclidean distance and cosine distance are supported.
203 """Computes cosine distance between each input and each cluster center.
210 cosine distance. Else it L2 normalizes the inputs first.
248 # The cosine distance between normalized vectors x and y is the same as
251 # TODO(ands): Support COSINE distance in nearest_neighbors and remove
553 l2-normalize new_centers if using cosine distance
/aosp_15_r20/packages/modules/Bluetooth/system/embdrv/sbc/decoder/srce/
Dsynthesis-sbc.c34 can be decomposed into the product of a type II discrete cosine kernel
43 cosine kernel given by
60 C2 * R corresponds to computing the discrete cosine transform of R, so
66 Wen-Chieh. "A Unified Fast Algorithm for Cosine Modulated
422 /* These values should be zero, since out[2] of the 4-band cosine modulation in SynthWindow40_int32_int32_symmetry_with_sum()
482 This routine implements the cosine modulation matrix for 4-subband
484 matrix, M4, can be factored into an 8-point Type II Discrete Cosine
501 (DCTII_4 * in) is computed using a Fast Cosine Transform. The algorithm
/aosp_15_r20/prebuilts/gcc/linux-x86/host/x86_64-w64-mingw32-4.8/share/info/
Dlibquadmath.info109 'acosq': arc cosine function
110 'acoshq': inverse hyperbolic cosine function
119 'coshq': hyperbolic cosine function
120 'cosq': cosine function
163 'sincosq': calculate sine and cosine simulataneously
178 'cacoshq': complex arc hyperbolic cosine function
179 'cacosq': complex arc cosine function
184 'ccosq' complex cosine function:
185 'ccoshq': complex hyperbolic cosine function
/aosp_15_r20/external/icing/icing/scoring/advanced_scoring/
H A Dadvanced-scorer_test.cc1370 [SearchSpecProto::EmbeddingQueryMetricType::COSINE] in TEST_F()
1415 "getEmbeddingParameter(0), \"COSINE\", 0))"), in TEST_F()
1471 [SearchSpecProto::EmbeddingQueryMetricType::COSINE] in TEST_F()
1498 "getEmbeddingParameter(1), \"COSINE\"))"), in TEST_F()
1506 HasSubstr("embedding query index 1 with metric type COSINE " in TEST_F()
1581 // COSINE: in TEST_F()
1592 .result_scores[0][SearchSpecProto::EmbeddingQueryMetricType::COSINE]; in TEST_F()
1648 "getEmbeddingParameter(0), \"COSINE\"))"), in TEST_F()
1663 ", \"COSINE\")) + " in TEST_F()
1694 // The second query does not contain cosine scores. in TEST_F()
[all …]
/aosp_15_r20/prebuilts/go/linux-x86/src/math/
Dsin.go8 Floating-point sine and cosine.
33 // Between pi/4 and pi/2 the cosine is represented as
49 // Circular cosine
62 // Between 0 and pi/4 the cosine is approximated by
112 // Cos returns the cosine of the radian argument x.
/aosp_15_r20/external/apache-commons-math/src/main/java/org/apache/commons/math3/transform/
H A DFastCosineTransformer.java30 * Implements the Fast Cosine Transform for transformation of one-dimensional real data sets. For
33 * <p>There are several variants of the discrete cosine transform. The present implementation
38 * if x<sub>0</sub>, &hellip;, x<sub>N-1</sub> is the data set to be cosine transformed, the
54 * <p>The present implementation of the discrete cosine transform as a fast cosine transform

12345678910>>...31