ultralytics-inference 0.0.56 Shape right-to-left and complex-script labels in native and web, with up to 2.3x faster label drawing

Cargo.toml

@@ -9,7 +9,7 @@ default-members = ["."]
 
 [package]
 name = "ultralytics-inference"
-version = "0.0.55"
+version = "0.0.56"
 edition = "2024"
 rust-version = "1.92"
 authors = [

README.md

@@ -244,7 +244,7 @@ ultralytics-inference predict
 ```text
 WARNING ⚠️ 'model' argument is missing. Using default '--model=yolo26n.onnx'.
 WARNING ⚠️ 'source' argument is missing. Using default images: https://ultralytics.com/images/bus.jpg, https://ultralytics.com/images/zidane.jpg
-Ultralytics Inference 0.0.55 🚀 Rust ONNX FP32 CPU
+Ultralytics Inference 0.0.56 🚀 Rust ONNX FP32 CPU
 Using ONNX Runtime CPUExecutionProvider
 YOLO26n summary: 80 classes, imgsz=(640, 640)
 
@@ -264,7 +264,7 @@ ultralytics-inference predict --task segment
 ```text
 WARNING ⚠️ 'model' argument is missing. Using default '--model=yolo26n-seg.onnx'.
 WARNING ⚠️ 'source' argument is missing. Using default images: https://ultralytics.com/images/bus.jpg, https://ultralytics.com/images/zidane.jpg
-Ultralytics Inference 0.0.55 🚀 Rust ONNX FP32 CPU
+Ultralytics Inference 0.0.56 🚀 Rust ONNX FP32 CPU
 Using ONNX Runtime CPUExecutionProvider
 YOLO26n-seg summary: 80 classes, imgsz=(640, 640)
 
@@ -371,7 +371,7 @@ Add to your `Cargo.toml` (choose one):
 ```toml
 # Stable release from crates.io
 [dependencies]
-ultralytics-inference = "0.0.55"
+ultralytics-inference = "0.0.56"
 ```
 
 ```toml
@@ -611,7 +611,7 @@ Each accelerator feature links a prebuilt ONNX Runtime containing that provider.
 
 ```toml
 [dependencies]
-ultralytics-inference = { version = "0.0.55", features = ["coreml", "xnnpack"] }
+ultralytics-inference = { version = "0.0.56", features = ["coreml", "xnnpack"] }
 ort = { version = "=2.0.0-rc.15", features = ["lax-feature-matching"] }
 ```
 
@@ -717,10 +717,10 @@ One of the key benefits of this library is a Rust/ONNX Runtime stack with no PyT
 
 ### Optional Dependencies (for the `annotate` feature)
 
-| Crate       | Purpose                                              |
-| ----------- | ---------------------------------------------------- |
-| `imageproc` | Drawing boxes and shapes                             |
-| `ab_glyph`  | Text rendering (Arial font, downloaded on first use) |
+| Crate       | Purpose                                                                                            |
+| ----------- | -------------------------------------------------------------------------------------------------- |
+| `imageproc` | Drawing boxes and shapes                                                                           |
+| `ab_glyph`  | Text rendering (Arial font, downloaded on first use; Arabic and Hebrew labels drawn right to left) |
 
 ### Optional Dependencies (for Video & Visualization)
 

README.zh-CN.md

@@ -243,7 +243,7 @@ ultralytics-inference predict
 ```text
 WARNING ⚠️ 'model' argument is missing. Using default '--model=yolo26n.onnx'.
 WARNING ⚠️ 'source' argument is missing. Using default images: https://ultralytics.com/images/bus.jpg, https://ultralytics.com/images/zidane.jpg
-Ultralytics Inference 0.0.55 🚀 Rust ONNX FP32 CPU
+Ultralytics Inference 0.0.56 🚀 Rust ONNX FP32 CPU
 Using ONNX Runtime CPUExecutionProvider
 YOLO26n summary: 80 classes, imgsz=(640, 640)
 
@@ -263,7 +263,7 @@ ultralytics-inference predict --task segment
 ```text
 WARNING ⚠️ 'model' argument is missing. Using default '--model=yolo26n-seg.onnx'.
 WARNING ⚠️ 'source' argument is missing. Using default images: https://ultralytics.com/images/bus.jpg, https://ultralytics.com/images/zidane.jpg
-Ultralytics Inference 0.0.55 🚀 Rust ONNX FP32 CPU
+Ultralytics Inference 0.0.56 🚀 Rust ONNX FP32 CPU
 Using ONNX Runtime CPUExecutionProvider
 YOLO26n-seg summary: 80 classes, imgsz=(640, 640)
 
@@ -370,7 +370,7 @@ YOLOv8、YOLO11 和 YOLO26 ONNX 模型支持 **n / s / m / l / x** 尺寸，并
 ```toml
 # crates.io 稳定版本
 [dependencies]
-ultralytics-inference = "0.0.55"
+ultralytics-inference = "0.0.56"
 ```
 
 ```toml
@@ -606,7 +606,7 @@ cargo build --release --features "cuda,tensorrt"
 
 ```toml
 [dependencies]
-ultralytics-inference = { version = "0.0.55", features = ["coreml", "xnnpack"] }
+ultralytics-inference = { version = "0.0.56", features = ["coreml", "xnnpack"] }
 ort = { version = "=2.0.0-rc.15", features = ["lax-feature-matching"] }
 ```
 
@@ -709,10 +709,10 @@ JS/TS 封装与构建说明见 [`web/`](web/README.md)。需要支持 WebGPU 的
 
 ### 可选依赖（用于 `annotate` feature）
 
-| Crate       | 用途                                   |
-| ----------- | -------------------------------------- |
-| `imageproc` | 绘制框和形状                           |
-| `ab_glyph`  | 文本渲染（Arial 字体，首次使用时下载） |
+| Crate       | 用途                                                                       |
+| ----------- | -------------------------------------------------------------------------- |
+| `imageproc` | 绘制框和形状                                                               |
+| `ab_glyph`  | 文本渲染（Arial 字体，首次使用时下载；阿拉伯语和希伯来语标签从右到左绘制） |
 
 ### 可选依赖（用于视频和可视化）
 

crates/web/Cargo.toml

@@ -2,7 +2,7 @@
 
 [package]
 name = "ultralytics-inference-web"
-version = "0.0.55"
+version = "0.0.56"
 edition = "2024"
 rust-version = "1.92"
 description = "Browser/WebAssembly bindings for ultralytics-inference (WebGPU via ort-web)"

docs/CUDA.md

@@ -39,9 +39,9 @@ Add the feature you need in your `Cargo.toml`:
 
 ```toml
 [dependencies]
-ultralytics-inference = { version = "0.0.55", features = ["tensorrt"] }
+ultralytics-inference = { version = "0.0.56", features = ["tensorrt"] }
 # or, for the fastest path:
-ultralytics-inference = { version = "0.0.55", features = ["cuda-preprocess"] }
+ultralytics-inference = { version = "0.0.56", features = ["cuda-preprocess"] }
 ```
 
 Then `cargo build --release`: no extra flags needed.
@@ -71,7 +71,7 @@ If you need to pin a specific version at compile time instead, override the cuda
 
 ```toml
 [dependencies]
-ultralytics-inference = { version = "0.0.55", features = ["cuda-preprocess"] }
+ultralytics-inference = { version = "0.0.56", features = ["cuda-preprocess"] }
 # Replace the default feature with a pinned one (e.g. CUDA 12.8):
 cudarc = { version = "0.19", default-features = false, features = ["driver", "nvrtc", "dynamic-loading", "cuda-12080"] }
 ```

docs/OPENVINO.md

@@ -7,7 +7,7 @@
 ```toml
 # Cargo.toml
 [dependencies]
-ultralytics-inference = { version = "0.0.55", features = ["openvino-native"] }
+ultralytics-inference = { version = "0.0.56", features = ["openvino-native"] }
 ```
 
 ```bash