Fixes multiple typos in the calculator's internal files.

PiperOrigin-RevId: 580907788

mediapipe/calculators/audio/spectrogram_calculator.proto

@@ -80,7 +80,7 @@ message SpectrogramCalculatorOptions {
   // If use_local_timestamp is true, the output packet's timestamp is based on
   // the last sample of the packet and it's inferred from the latest input
   // packet's timestamp.  If false, the output packet's timestamp is based on
-  // the cumulative timestamping, which is inferred from the intial input
+  // the cumulative timestamping, which is inferred from the initial input
   // timestamp and the cumulative number of samples.
   optional bool use_local_timestamp = 8 [default = false];
 }

mediapipe/calculators/audio/time_series_framer_calculator.proto

@@ -66,7 +66,7 @@ message TimeSeriesFramerCalculatorOptions {
   // If use_local_timestamp is true, the output packet's timestamp is based on
   // the last sample of the packet and it's inferred from the latest input
   // packet's timestamp.  If false, the output packet's timestamp is based on
-  // the cumulative timestamping, which is inferred from the intial input
+  // the cumulative timestamping, which is inferred from the initial input
   // timestamp and the cumulative number of samples.
   optional bool use_local_timestamp = 6 [default = false];
 }

mediapipe/calculators/core/packet_sequencer_calculator_test.cc

@@ -71,7 +71,7 @@ TEST_F(PacketSequencerCalculatorTest, IsRegistered) {
       CalculatorBaseRegistry::IsRegistered("PacketSequencerCalculator"));
 }
 
-// Shows how control packets recieve timestamps before and after frame packets
+// Shows how control packets receive timestamps before and after frame packets
 // have arrived.
 TEST_F(PacketSequencerCalculatorTest, ChannelEarly) {
   CalculatorGraphConfig::Node node_config = BuildNodeConfig();

mediapipe/calculators/core/value_or_default_calculator_test.cc

@@ -174,7 +174,7 @@ TEST(ValueOrDefaultCalculatorTest, DefaultAndValues) {
       ElementsAre(kDefaultValue, 1, 2, kDefaultValue, 3, kDefaultValue));
 }
 
-TEST(ValueOrDefaultCalculatorTest, TimestampsMissmatch) {
+TEST(ValueOrDefaultCalculatorTest, TimestampsMismatch) {
   // Check that when we provide the inputs not on time - we don't get them.
   ValueOrDefaultRunner runner;
   const std::vector<int64_t> ticks = {1, 2, 5, 8, 12, 33, 231};

mediapipe/calculators/image/affine_transformation_runner_opencv.cc

@@ -59,7 +59,7 @@ class OpenCvRunner
       const ImageFrame& input, const std::array<float, 16>& matrix,
       const AffineTransformation::Size& size,
       AffineTransformation::BorderMode border_mode) override {
-    // OpenCV warpAffine works in absolute coordinates, so the transfom (which
+    // OpenCV warpAffine works in absolute coordinates, so the transform (which
     // accepts and produces relative coordinates) should be adjusted to first
     // normalize coordinates and then scale them.
     // clang-format off

mediapipe/calculators/image/image_cropping_calculator.proto

@@ -24,7 +24,7 @@ message ImageCroppingCalculatorOptions {
   }
 
   // Output texture buffer dimensions. The values defined in the options will be
-  // overriden by the WIDTH and HEIGHT input streams if they exist.
+  // overridden by the WIDTH and HEIGHT input streams if they exist.
   optional int32 width = 1;
   optional int32 height = 2;
 

mediapipe/calculators/image/image_file_properties_calculator.cc

@@ -77,7 +77,7 @@ absl::StatusOr<double> ComputeFocalLengthInPixels(int image_width,
   return focal_length_pixels;
 }
 
-absl::StatusOr<ImageFileProperties> GetImageFileProperites(
+absl::StatusOr<ImageFileProperties> GetImageFileProperties(
     const std::string& image_bytes) {
   easyexif::EXIFInfo result;
   int code = result.parseFrom(image_bytes);
@@ -151,7 +151,7 @@ class ImageFilePropertiesCalculator : public CalculatorBase {
     if (cc->InputSidePackets().NumEntries() == 1) {
       const std::string& image_bytes =
           cc->InputSidePackets().Index(0).Get<std::string>();
-      MP_ASSIGN_OR_RETURN(properties_, GetImageFileProperites(image_bytes));
+      MP_ASSIGN_OR_RETURN(properties_, GetImageFileProperties(image_bytes));
       read_properties_ = true;
     }
 
@@ -169,7 +169,7 @@ class ImageFilePropertiesCalculator : public CalculatorBase {
         return absl::OkStatus();
       }
       const std::string& image_bytes = cc->Inputs().Index(0).Get<std::string>();
-      MP_ASSIGN_OR_RETURN(properties_, GetImageFileProperites(image_bytes));
+      MP_ASSIGN_OR_RETURN(properties_, GetImageFileProperties(image_bytes));
       read_properties_ = true;
     }
     if (read_properties_) {

mediapipe/calculators/image/warp_affine_calculator_test.cc

@@ -284,7 +284,7 @@ std::array<float, 16> GetMatrix(cv::Mat input, mediapipe::NormalizedRect roi,
       .IgnoreError();
   mediapipe::GetRotatedSubRectToRectTransformMatrix(
       roi_absolute, input.cols, input.rows,
-      /*flip_horizontaly=*/false, &transform_mat);
+      /*flip_horizontally=*/false, &transform_mat);
   return transform_mat;
 }
 

mediapipe/calculators/image/yuv_to_image_calculator.cc

@@ -49,7 +49,7 @@ std::string FourCCToString(libyuv::FourCC fourcc) {
 // The input `YUVImage` is expected to be in the NV12, NV21, YV12 or I420 (aka
 // YV21) format (as per the `fourcc()` property). This covers the most commonly
 // used YUV image formats used on mobile devices. Other formats are not
-// supported and wil result in an `InvalidArgumentError`.
+// supported and will result in an `InvalidArgumentError`.
 class YUVToImageCalculator : public Node {
  public:
   static constexpr Input<YUVImage> kInput{"YUV_IMAGE"};

mediapipe/calculators/tflite/testdata/README.md

@@ -1,2 +1,2 @@
 The model files add.bin, add_quantized.bin
-(and corresponding metatada json files) come from tensorflow/lite/testdata/
+(and corresponding metadata json files) come from tensorflow/lite/testdata/

mediapipe/calculators/tflite/tflite_converter_calculator.cc

@@ -95,7 +95,7 @@ struct GPUData {
 // into a TfLiteTensor (float 32) or a GpuBuffer to a tflite::gpu::GlBuffer
 // or MTLBuffer.
 //
-// This calculator is designed to be used with the TfLiteInferenceCalcualtor,
+// This calculator is designed to be used with the TfLiteInferenceCalculator,
 // as a pre-processing step for calculator inputs.
 //
 // IMAGE and IMAGE_GPU inputs are normalized to [-1,1] (default) or [0,1],

mediapipe/calculators/tflite/tflite_converter_calculator.proto

@@ -31,7 +31,7 @@ message TfLiteConverterCalculatorOptions {
   // Custom settings to override the internal scaling factors `div` and `sub`.
   // Both values must be set to non-negative values. Will only take effect on
   // CPU AND when |use_custom_normalization| is set to true. When these custom
-  // values take effect, the |zero_center| setting above will be overriden, and
+  // values take effect, the |zero_center| setting above will be overridden, and
   // the normalized_value will be calculated as:
   // normalized_value = input / custom_div - custom_sub.
   optional bool use_custom_normalization = 6 [default = false];

mediapipe/calculators/tflite/tflite_tensors_to_classification_calculator.proto

@@ -25,7 +25,7 @@ message TfLiteTensorsToClassificationCalculatorOptions {
     optional TfLiteTensorsToClassificationCalculatorOptions ext = 266399463;
   }
 
-  // Score threshold for perserving the class.
+  // Score threshold for preserving the class.
   optional float min_score_threshold = 1;
   // Number of highest scoring labels to output.  If top_k is not positive then
   // all labels are used.

mediapipe/calculators/tflite/tflite_tensors_to_detections_calculator.cc

@@ -116,7 +116,7 @@ void ConvertAnchorsToRawValues(const std::vector<Anchor>& anchors,
 //               tensors can have 2 or 3 tensors. First tensor is the predicted
 //               raw boxes/keypoints. The size of the values must be (num_boxes
 //               * num_predicted_values). Second tensor is the score tensor. The
-//               size of the valuse must be (num_boxes * num_classes). It's
+//               size of the values must be (num_boxes * num_classes). It's
 //               optional to pass in a third tensor for anchors (e.g. for SSD
 //               models) depend on the outputs of the detection model. The size
 //               of anchor tensor must be (num_boxes * 4).

mediapipe/calculators/tflite/tflite_tensors_to_detections_calculator.proto

@@ -69,6 +69,6 @@ message TfLiteTensorsToDetectionsCalculatorOptions {
   // representation has a bottom-left origin (e.g., in OpenGL).
   optional bool flip_vertically = 18 [default = false];
 
-  // Score threshold for perserving decoded detections.
+  // Score threshold for preserving decoded detections.
   optional float min_score_thresh = 19;
 }

mediapipe/calculators/tflite/tflite_tensors_to_landmarks_calculator.cc

@@ -158,7 +158,7 @@ absl::Status TfLiteTensorsToLandmarksCalculator::Open(CalculatorContext* cc) {
     RET_CHECK(options_.has_input_image_height() &&
               options_.has_input_image_width())
         << "Must provide input width/height for using flip_vertically option "
-           "when outputing landmarks in absolute coordinates.";
+           "when outputting landmarks in absolute coordinates.";
   }
 
   flip_horizontally_ =