Reference for ultralytics/models/yolo/segment/val.py

def _prepare_batch(self, si, batch):
    """
    Prepare a batch for training or inference by processing images and targets.

    Args:
        si (int): Batch index.
        batch (Dict): Batch data containing images and targets.

    Returns:
        (Dict): Prepared batch with processed images and targets.
    """
    prepared_batch = super()._prepare_batch(si, batch)
    midx = [si] if self.args.overlap_mask else batch["batch_idx"] == si
    prepared_batch["masks"] = batch["masks"][midx]
    return prepared_batch

def _prepare_pred(self, pred, pbatch, proto):
    """
    Prepare predictions for evaluation by processing bounding boxes and masks.

    Args:
        pred (torch.Tensor): Raw predictions from the model.
        pbatch (Dict): Prepared batch data.
        proto (torch.Tensor): Prototype masks for segmentation.

    Returns:
        predn (torch.Tensor): Processed bounding box predictions.
        pred_masks (torch.Tensor): Processed mask predictions.
    """
    predn = super()._prepare_pred(pred, pbatch)
    pred_masks = self.process(proto, pred[:, 6:], pred[:, :4], shape=pbatch["imgsz"])
    return predn, pred_masks

def _process_batch(self, detections, gt_bboxes, gt_cls, pred_masks=None, gt_masks=None, overlap=False, masks=False):
    """
    Compute correct prediction matrix for a batch based on bounding boxes and optional masks.

    Args:
        detections (torch.Tensor): Tensor of shape (N, 6) representing detected bounding boxes and
            associated confidence scores and class indices. Each row is of the format [x1, y1, x2, y2, conf, class].
        gt_bboxes (torch.Tensor): Tensor of shape (M, 4) representing ground truth bounding box coordinates.
            Each row is of the format [x1, y1, x2, y2].
        gt_cls (torch.Tensor): Tensor of shape (M,) representing ground truth class indices.
        pred_masks (torch.Tensor, optional): Tensor representing predicted masks, if available. The shape should
            match the ground truth masks.
        gt_masks (torch.Tensor, optional): Tensor of shape (M, H, W) representing ground truth masks, if available.
        overlap (bool): Flag indicating if overlapping masks should be considered.
        masks (bool): Flag indicating if the batch contains mask data.

    Returns:
        (torch.Tensor): A correct prediction matrix of shape (N, 10), where 10 represents different IoU levels.

    Note:
        - If `masks` is True, the function computes IoU between predicted and ground truth masks.
        - If `overlap` is True and `masks` is True, overlapping masks are taken into account when computing IoU.

    Examples:
        >>> detections = torch.tensor([[25, 30, 200, 300, 0.8, 1], [50, 60, 180, 290, 0.75, 0]])
        >>> gt_bboxes = torch.tensor([[24, 29, 199, 299], [55, 65, 185, 295]])
        >>> gt_cls = torch.tensor([1, 0])
        >>> correct_preds = validator._process_batch(detections, gt_bboxes, gt_cls)
    """
    if masks:
        if overlap:
            nl = len(gt_cls)
            index = torch.arange(nl, device=gt_masks.device).view(nl, 1, 1) + 1
            gt_masks = gt_masks.repeat(nl, 1, 1)  # shape(1,640,640) -> (n,640,640)
            gt_masks = torch.where(gt_masks == index, 1.0, 0.0)
        if gt_masks.shape[1:] != pred_masks.shape[1:]:
            gt_masks = F.interpolate(gt_masks[None], pred_masks.shape[1:], mode="bilinear", align_corners=False)[0]
            gt_masks = gt_masks.gt_(0.5)
        iou = mask_iou(gt_masks.view(gt_masks.shape[0], -1), pred_masks.view(pred_masks.shape[0], -1))
    else:  # boxes
        iou = box_iou(gt_bboxes, detections[:, :4])

    return self.match_predictions(detections[:, 5], gt_cls, iou)

def eval_json(self, stats):
    """Return COCO-style object detection evaluation metrics."""
    if self.args.save_json and self.is_coco and len(self.jdict):
        anno_json = self.data["path"] / "annotations/instances_val2017.json"  # annotations
        pred_json = self.save_dir / "predictions.json"  # predictions
        LOGGER.info(f"\nEvaluating pycocotools mAP using {pred_json} and {anno_json}...")
        try:  # https://github.com/cocodataset/cocoapi/blob/master/PythonAPI/pycocoEvalDemo.ipynb
            check_requirements("pycocotools>=2.0.6")
            from pycocotools.coco import COCO  # noqa
            from pycocotools.cocoeval import COCOeval  # noqa

            for x in anno_json, pred_json:
                assert x.is_file(), f"{x} file not found"
            anno = COCO(str(anno_json))  # init annotations api
            pred = anno.loadRes(str(pred_json))  # init predictions api (must pass string, not Path)
            for i, eval in enumerate([COCOeval(anno, pred, "bbox"), COCOeval(anno, pred, "segm")]):
                if self.is_coco:
                    eval.params.imgIds = [int(Path(x).stem) for x in self.dataloader.dataset.im_files]  # im to eval
                eval.evaluate()
                eval.accumulate()
                eval.summarize()
                idx = i * 4 + 2
                stats[self.metrics.keys[idx + 1]], stats[self.metrics.keys[idx]] = eval.stats[
                    :2
                ]  # update mAP50-95 and mAP50
        except Exception as e:
            LOGGER.warning(f"pycocotools unable to run: {e}")
    return stats

def finalize_metrics(self, *args, **kwargs):
    """Set speed and confusion matrix for evaluation metrics."""
    self.metrics.speed = self.speed
    self.metrics.confusion_matrix = self.confusion_matrix

def get_desc(self):
    """Return a formatted description of evaluation metrics."""
    return ("%22s" + "%11s" * 10) % (
        "Class",
        "Images",
        "Instances",
        "Box(P",
        "R",
        "mAP50",
        "mAP50-95)",
        "Mask(P",
        "R",
        "mAP50",
        "mAP50-95)",
    )

def init_metrics(self, model):
    """
    Initialize metrics and select mask processing function based on save_json flag.

    Args:
        model (torch.nn.Module): Model to validate.
    """
    super().init_metrics(model)
    self.plot_masks = []
    if self.args.save_json:
        check_requirements("pycocotools>=2.0.6")
    # more accurate vs faster
    self.process = ops.process_mask_native if self.args.save_json or self.args.save_txt else ops.process_mask
    self.stats = dict(tp_m=[], tp=[], conf=[], pred_cls=[], target_cls=[], target_img=[])

def plot_predictions(self, batch, preds, ni):
    """
    Plot batch predictions with masks and bounding boxes.

    Args:
        batch (Dict): Batch data containing images.
        preds (List): Predictions from the model.
        ni (int): Batch index.
    """
    plot_images(
        batch["img"],
        *output_to_target(preds[0], max_det=15),  # not set to self.args.max_det due to slow plotting speed
        torch.cat(self.plot_masks, dim=0) if len(self.plot_masks) else self.plot_masks,
        paths=batch["im_file"],
        fname=self.save_dir / f"val_batch{ni}_pred.jpg",
        names=self.names,
        on_plot=self.on_plot,
    )  # pred
    self.plot_masks.clear()

def plot_val_samples(self, batch, ni):
    """
    Plot validation samples with bounding box labels and masks.

    Args:
        batch (Dict): Batch data containing images and targets.
        ni (int): Batch index.
    """
    plot_images(
        batch["img"],
        batch["batch_idx"],
        batch["cls"].squeeze(-1),
        batch["bboxes"],
        masks=batch["masks"],
        paths=batch["im_file"],
        fname=self.save_dir / f"val_batch{ni}_labels.jpg",
        names=self.names,
        on_plot=self.on_plot,
    )

def postprocess(self, preds):
    """
    Post-process YOLO predictions and return output detections with proto.

    Args:
        preds (List): Raw predictions from the model.

    Returns:
        p (torch.Tensor): Processed detection predictions.
        proto (torch.Tensor): Prototype masks for segmentation.
    """
    p = super().postprocess(preds[0])
    proto = preds[1][-1] if len(preds[1]) == 3 else preds[1]  # second output is len 3 if pt, but only 1 if exported
    return p, proto

def pred_to_json(self, predn, filename, pred_masks):
    """
    Save one JSON result for COCO evaluation.

    Args:
        predn (torch.Tensor): Predictions in the format [x1, y1, x2, y2, conf, cls].
        filename (str): Image filename.
        pred_masks (numpy.ndarray): Predicted masks.

    Examples:
         >>> result = {"image_id": 42, "category_id": 18, "bbox": [258.15, 41.29, 348.26, 243.78], "score": 0.236}
    """
    from pycocotools.mask import encode  # noqa

    def single_encode(x):
        """Encode predicted masks as RLE and append results to jdict."""
        rle = encode(np.asarray(x[:, :, None], order="F", dtype="uint8"))[0]
        rle["counts"] = rle["counts"].decode("utf-8")
        return rle

    stem = Path(filename).stem
    image_id = int(stem) if stem.isnumeric() else stem
    box = ops.xyxy2xywh(predn[:, :4])  # xywh
    box[:, :2] -= box[:, 2:] / 2  # xy center to top-left corner
    pred_masks = np.transpose(pred_masks, (2, 0, 1))
    with ThreadPool(NUM_THREADS) as pool:
        rles = pool.map(single_encode, pred_masks)
    for i, (p, b) in enumerate(zip(predn.tolist(), box.tolist())):
        self.jdict.append(
            {
                "image_id": image_id,
                "category_id": self.class_map[int(p[5])],
                "bbox": [round(x, 3) for x in b],
                "score": round(p[4], 5),
                "segmentation": rles[i],
            }
        )

def preprocess(self, batch):
    """Preprocess batch by converting masks to float and sending to device."""
    batch = super().preprocess(batch)
    batch["masks"] = batch["masks"].to(self.device).float()
    return batch

def save_one_txt(self, predn, pred_masks, save_conf, shape, file):
    """
    Save YOLO detections to a txt file in normalized coordinates in a specific format.

    Args:
        predn (torch.Tensor): Predictions in the format [x1, y1, x2, y2, conf, cls].
        pred_masks (torch.Tensor): Predicted masks.
        save_conf (bool): Whether to save confidence scores.
        shape (Tuple): Original image shape.
        file (Path): File path to save the detections.
    """
    from ultralytics.engine.results import Results

    Results(
        np.zeros((shape[0], shape[1]), dtype=np.uint8),
        path=None,
        names=self.names,
        boxes=predn[:, :6],
        masks=pred_masks,
    ).save_txt(file, save_conf=save_conf)

def update_metrics(self, preds, batch):
    """
    Update metrics with the current batch predictions and targets.

    Args:
        preds (List): Predictions from the model.
        batch (Dict): Batch data containing images and targets.
    """
    for si, (pred, proto) in enumerate(zip(preds[0], preds[1])):
        self.seen += 1
        npr = len(pred)
        stat = dict(
            conf=torch.zeros(0, device=self.device),
            pred_cls=torch.zeros(0, device=self.device),
            tp=torch.zeros(npr, self.niou, dtype=torch.bool, device=self.device),
            tp_m=torch.zeros(npr, self.niou, dtype=torch.bool, device=self.device),
        )
        pbatch = self._prepare_batch(si, batch)
        cls, bbox = pbatch.pop("cls"), pbatch.pop("bbox")
        nl = len(cls)
        stat["target_cls"] = cls
        stat["target_img"] = cls.unique()
        if npr == 0:
            if nl:
                for k in self.stats.keys():
                    self.stats[k].append(stat[k])
                if self.args.plots:
                    self.confusion_matrix.process_batch(detections=None, gt_bboxes=bbox, gt_cls=cls)
            continue

        # Masks
        gt_masks = pbatch.pop("masks")
        # Predictions
        if self.args.single_cls:
            pred[:, 5] = 0
        predn, pred_masks = self._prepare_pred(pred, pbatch, proto)
        stat["conf"] = predn[:, 4]
        stat["pred_cls"] = predn[:, 5]

        # Evaluate
        if nl:
            stat["tp"] = self._process_batch(predn, bbox, cls)
            stat["tp_m"] = self._process_batch(
                predn, bbox, cls, pred_masks, gt_masks, self.args.overlap_mask, masks=True
            )
        if self.args.plots:
            self.confusion_matrix.process_batch(predn, bbox, cls)

        for k in self.stats.keys():
            self.stats[k].append(stat[k])

        pred_masks = torch.as_tensor(pred_masks, dtype=torch.uint8)
        if self.args.plots and self.batch_i < 3:
            self.plot_masks.append(pred_masks[:15].cpu())  # filter top 15 to plot

        # Save
        if self.args.save_json:
            self.pred_to_json(
                predn,
                batch["im_file"][si],
                ops.scale_image(
                    pred_masks.permute(1, 2, 0).contiguous().cpu().numpy(),
                    pbatch["ori_shape"],
                    ratio_pad=batch["ratio_pad"][si],
                ),
            )
        if self.args.save_txt:
            self.save_one_txt(
                predn,
                pred_masks,
                self.args.save_conf,
                pbatch["ori_shape"],
                self.save_dir / "labels" / f"{Path(batch['im_file'][si]).stem}.txt",
            )