# Miscalibrated yolo26

**URL:** <https://community.ultralytics.com/t/miscalibrated-yolo26/1847>\
**Category:** Discussion\
**Tags:** discussion, yolo\
**Created:** [March 6, 2026, 8:32pm UTC](https://community.ultralytics.com/t/miscalibrated-yolo26/1847 "2026-03-06T20:32:39Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![giovanni-dd](https://sea1.discourse-cdn.com/flex001/user_avatar/community.ultralytics.com/giovanni-dd/32/1354_2.png) [@giovanni-dd](https://community.ultralytics.com/u/giovanni-dd)\
**Post date:** [March 6, 2026, 8:32pm UTC](https://community.ultralytics.com/t/miscalibrated-yolo26/1847/1 "2026-03-06T20:32:39Z")

</div>

Hello Everyone! I trained a yolo26 model to detect tiny objects, performance is decent-ish, with map@50 0.62, in line with other architectures I tried, the problem is the model has very low confidence, with maximum f1 archieved at p=0.003, making even the map calculation numerically questionable.  
This is probably due to class imbalance, most image is background, with a tiny target somewhere, some images have no target at all.  
With older models I could tune the parameter alpha in the focal loss to address class imbalance. Is there some similar parameter in the more involved yolo26 loss to weight less the true negatives? I am detecting single class if it is relevant.

---

<div class="post-metadata">

**Author:** ![pderrenger](https://sea1.discourse-cdn.com/flex001/user_avatar/community.ultralytics.com/pderrenger/32/73_2.png) [@pderrenger](https://community.ultralytics.com/u/pderrenger)\
**Post date:** [March 7, 2026, 1:30pm UTC](https://community.ultralytics.com/t/miscalibrated-yolo26/1847/2 "2026-03-07T13:30:56Z")

</div>

With extreme foreground/background imbalance it’s pretty normal for the optimal F1 to land at a very low `conf` (your model is basically saying “almost everything is background”), and that **doesn’t make mAP invalid** —mAP is based on ranking predictions by score across thresholds, not on any single “reasonable” confidence value. For inference you can simply run with `conf=0.003` (or whatever your PR/F1 curve suggests).

If you _want_ to explicitly reduce the impact of true negatives / boost positives like you did with focal-loss `alpha`, Ultralytics YOLO26 doesn’t expose a simple “alpha knob” in the CLI, but you can do the equivalent by **upweighting the positive class in the BCE classification loss** (per-class `pos_weight`) via a custom trainer. The docs show the exact pattern here in the [Customizing Trainer guide (Adding Class Weights)](https://docs.ultralytics.com/guides/custom-trainer/).

For a single-class dataset, it’s as simple as setting a 1-element weight tensor, e.g. `torch.tensor([10.0])`, in `init_criterion()` (adapted from the doc example):

```python
import torch
from torch import nn
from ultralytics import YOLO
from ultralytics.models.yolo.detect import DetectionTrainer
from ultralytics.nn.tasks import DetectionModel
from ultralytics.utils import RANK
from ultralytics.utils.loss import E2ELoss, v8DetectionLoss

class WeightedDetectionLoss(v8DetectionLoss):
    def __init__ (self, model, class_weights=None, tal_topk=10, tal_topk2=None):
        super(). __init__ (model, tal_topk=tal_topk, tal_topk2=tal_topk2)
        if class_weights is not None:
            self.bce = nn.BCEWithLogitsLoss(
                pos_weight=class_weights.to(self.device),
                reduction="none",
            )

class WeightedE2ELoss(E2ELoss):
    def __init__ (self, model, class_weights=None):
        super(). __init__ (model, loss_fn=lambda m, tal_topk=10, tal_topk2=None:
                         WeightedDetectionLoss(m, class_weights=class_weights, tal_topk=tal_topk, tal_topk2=tal_topk2))

class WeightedDetectionModel(DetectionModel):
    def init_criterion(self):
        # single-class: shape [1]
        class_weights = torch.tensor([10.0])
        return WeightedE2ELoss(self, class_weights=class_weights)

class WeightedTrainer(DetectionTrainer):
    def get_model(self, cfg=None, weights=None, verbose=True):
        model = WeightedDetectionModel(cfg, nc=self.data["nc"], verbose=verbose and RANK == -1)
        if weights:
            model.load(weights)
        return model

model = YOLO("yolo26n.pt")
model.train(data="your.yaml", epochs=100, trainer=WeightedTrainer)

```

If you share your label stats (how many images are empty vs contain an object, plus object size distribution and `imgsz`), I can suggest a sane starting weight and a couple of training tweaks for tiny targets.

---

<div class="post-metadata">

**Author:** ![giovanni-dd](https://sea1.discourse-cdn.com/flex001/user_avatar/community.ultralytics.com/giovanni-dd/32/1354_2.png) [@giovanni-dd](https://community.ultralytics.com/u/giovanni-dd)\
**Post date:** [March 11, 2026, 1:14pm UTC](https://community.ultralytics.com/t/miscalibrated-yolo26/1847/3 "2026-03-11T13:14:43Z")

</div>

Thank you for the detailed answer! Will try it.  
My training set has 640x512 images, half with a target, half empty.  
Most (but not all) of the annotations are small bright objects \< 100px in area

 ![label_distro](https://us1.discourse-cdn.com/flex001/uploads/ultralytics1/original/2X/1/1e63dc85cdbc23fafaa1a7c1e942c40c293858a8.png)