Skip to content

Getting Started⚓︎

This guide will walk you through installing NextCV and running your first performance demo.

1. Prerequisites⚓︎

sudo apt-get install libeigen3-dev cmake

brew install eigen cmake

2. Installation⚓︎

pip install git+https://github.com/kevinconka/nextcv.git

uv add git+https://github.com/kevinconka/nextcv.git

3. Performance Demo⚓︎

This example demonstrates the performance difference between the C++ and NumPy implementations of Non-Maximum Suppression (NMS).

import time
import numpy as np
from nextcv.postprocessing import nms_cpp, nms_np

# Create a large dataset of bounding boxes
N = 10000
rng = np.random.default_rng(42)
bboxes = rng.uniform(0, 100, (N, 4)).astype(np.float32)
scores = rng.uniform(0.1, 1, N).astype(np.float32)

# Time C++ implementation
start_time = time.perf_counter()
result_cpp = nms_cpp(bboxes, scores, 0.5)
cpp_time = time.perf_counter() - start_time

# Time NumPy implementation
start_time = time.perf_counter()
result_np = nms_np(bboxes, scores, 0.5)
np_time = time.perf_counter() - start_time

print("Post-processing (NMS timing comparison):")
print(f"   Dataset: {len(bboxes)} bounding boxes")
print(f"   nms_cpp(): {len(result_cpp)} boxes kept in {cpp_time * 1000:.2f}ms")
print(f"   nms_np(): {len(result_np)} boxes kept in {np_time * 1000:.2f}ms")

Output:

Post-processing (NMS timing comparison):
   Dataset: 10000 bounding boxes
   nms_cpp(): 950 boxes kept in 25.93ms
   nms_np(): 949 boxes kept in 65.39ms

4. Next Steps⚓︎

Now that you have NextCV installed, you can start exploring the API.

  • Explore the API: Check out the API Reference for a full list of available functions.
  • Learn about our philosophy: Read our guide on When to Use C++ to understand our approach to performance.