Merge pull request #1 from kianmeng/fix-typos

Fix typos

README.md

@@ -67,7 +67,7 @@ By default, the app's settings are persisted after closing.
 The purpose of this crate is to study tradeoffs regarding model inference, native GUIs and
 video decoding approaches, in Rust :crab:.
 
-There are a couple of Todos will make `InFur` more intersting beyond exploring
+There are a couple of Todos will make `InFur` more interesting beyond exploring
 production-readiness as now:
 
 - [ ] GATify `type Output` in `trait Processor`

ff-video/src/decoder.rs

@@ -174,7 +174,7 @@ enum StreamInfoTerm {
     Final(String),
 }
 
-/// Deliver infos about an ffmpeg video process trhough its stderr file
+/// Deliver infos about an ffmpeg video process through its stderr file
 ///
 /// The receiver can be read until satisfying info was obtained and dropped anytime.
 /// By default, frame updates and other infos are logged as tracing event.

ff-video/src/parse.rs

@@ -251,7 +251,7 @@ impl InfoParser {
 
 /// Blanket implementation for lines of ffmpeg's default stderr bytes.
 pub(crate) trait FFMpegLineIter: Iterator {
-    /// Emit lines on \n, \r (CR) or both but never emtpy lines.
+    /// Emit lines on \n, \r (CR) or both but never empty lines.
     fn ffmpeg_lines(self) -> FFMpegLines<Self>
     where
         Self: Sized,

infur-test-gen/build.rs

@@ -27,7 +27,7 @@ fn run_ffmpeg_synth(
         .expect("synthesizing video couldn't start, do you have ffmpeg in PATH?")
         .wait()
         .expect("synthesizing video didn't finish");
-    assert!(status.success(), "synthesizing videos didn't finish succesfully");
+    assert!(status.success(), "synthesizing videos didn't finish successfully");
 }
 
 fn download(source_url: &str, target_file: impl AsRef<Path>) {
@@ -85,7 +85,7 @@ pub fn main() {
     }
 
     // models
-    // segementation model, see: https://github.com/onnx/models/tree/main/vision/object_detection_segmentation/fcn
+    // segmentation model, see: https://github.com/onnx/models/tree/main/vision/object_detection_segmentation/fcn
     let fcn_resnet50_12_int8 = gen_root.join("models").join("fcn-resnet50-12-int8.onnx");
     download("https://github.com/onnx/models/raw/main/vision/object_detection_segmentation/fcn/model/fcn-resnet50-12-int8.onnx",
         &fcn_resnet50_12_int8);

infur/Cargo.toml

@@ -23,7 +23,7 @@ eframe = { version = "0.19", features = ["wgpu", "default_fonts"], default-featu
 serde = { version = "1", features = ["derive"] }
 fast_image_resize = { version = "1" }
 # need onnxruntime .14 for 0-dim input tolerance (not in .13),
-# then furhtermore need master to resolve ndarray with tract-core...
+# then furthermore need master to resolve ndarray with tract-core...
 onnxruntime = { git = "https://github.com/nbigaouette/onnxruntime-rs" }
 once_cell = "1"
 image-ext = { path = "../image-ext" }

infur/src/decode_predict.rs

@@ -6,7 +6,7 @@ use onnxruntime::ndarray::Array3;
 ///
 /// adapted from: <http://www.color-hex.com/color-palette/23381>
 /// and: <http://www.color-hex.com/color-palette/52402>
-const COLORS_PALATTE: [(u8, u8, u8); 20] = [
+const COLORS_PALETTE: [(u8, u8, u8); 20] = [
     (75, 180, 60),
     (75, 25, 230),
     (25, 225, 255),
@@ -31,7 +31,7 @@ const COLORS_PALATTE: [(u8, u8, u8); 20] = [
 
 fn color_code(klass: usize, alpha: f32) -> Color32 {
     // todo: pre-transform COLORS into linear space
-    let (r, g, b) = COLORS_PALATTE[klass % COLORS_PALATTE.len()];
+    let (r, g, b) = COLORS_PALETTE[klass % COLORS_PALETTE.len()];
     Color32::from_rgba_unmultiplied(r, g, b, (alpha * 255.0f32) as u8)
 }
 
@@ -92,7 +92,7 @@ mod test {
 
     #[test]
     fn color_2() {
-        let c = COLORS_PALATTE[2];
+        let c = COLORS_PALETTE[2];
         assert_eq!(color_code(2, 0.5), Color32::from_rgba_unmultiplied(c.0, c.1, c.2, 127));
     }
 

infur/src/predict_onnx.rs

@@ -70,7 +70,7 @@ struct ImageSession<'s> {
 
 /// ONNX session with pre-processing u8 images.
 impl<'s> ImageSession<'s> {
-    /// Constract an `ImageSession` by inferring some required image input meta data.
+    /// Construct an `ImageSession` by inferring some required image input meta data.
     ///
     /// The basic assumption is that images are passed as batches at position 0.
     ///
@@ -375,7 +375,7 @@ mod test {
         let mut tensors = vec![];
         m.advance(&img, &mut tensors).unwrap();
 
-        assert_eq!(tensors.len(), 2, "this sementation model should return two tensors");
+        assert_eq!(tensors.len(), 2, "this segmentation model should return two tensors");
         assert_eq!(tensors[0].shape(), [21, 240, 320], "out should be 21 classes upscaled");
         assert_eq!(tensors[1].shape(), [21, 240, 320], "aux should be 21 classes upscaled");
     }