feat(trackers): Implement OC-SORT

[soumik12345]

Description

The PR implements the OC-SORT tracking algorithm from the paper Observation-Centric SORT: Rethinking SORT for Robust Multi-Object Tracking.

Type of change

• New feature (non-breaking change which adds functionality)

How has this change been tested, please provide a testcase or example of how you tested the change?

The code has been tested with the following code snippet:

import supervision as sv
from ultralytics import YOLO

from trackers import OCSORTTracker

model = YOLO("yolo11l.pt")
tracker = OCSORTTracker()

color = sv.ColorPalette.from_hex(
    [
        "#ffff00",
        "#ff9b00",
        "#ff8080",
        "#ff66b2",
        "#ff66ff",
        "#b266ff",
        "#9999ff",
        "#3399ff",
        "#66ffff",
        "#33ff99",
        "#66ff66",
        "#99ff00",
    ]
)

box_annotator = sv.BoxAnnotator(color=color, color_lookup=sv.ColorLookup.TRACK)

trace_annotator = sv.TraceAnnotator(
    color=color, color_lookup=sv.ColorLookup.TRACK, thickness=2, trace_length=100
)

label_annotator = sv.LabelAnnotator(
    color=color,
    color_lookup=sv.ColorLookup.TRACK,
    text_color=sv.Color.BLACK,
    text_scale=0.8,
)


def callback(frame, i):
    result = model(frame, imgsz=1280, verbose=False)[0]
    detections = sv.Detections.from_ultralytics(result)
    detections = tracker.update(detections)

    if not detections.is_empty():
        detections = detections[detections.class_id == 0]

    annotated_image = frame.copy()
    annotated_image = box_annotator.annotate(annotated_image, detections)
    annotated_image = trace_annotator.annotate(annotated_image, detections)
    annotated_image = label_annotator.annotate(
        annotated_image, detections, detections.tracker_id
    )

    return annotated_image


tracker.reset()

sv.process_video(
    source_path="people-walking.mp4",
    target_path="output.mp4",
    callback=callback,
    show_progress=True,
)

Docs

• [❌] Docs updated? What were the changes