Recall

Recall, also known as sensitivity or the true positive rate, is a fundamental performance metric in machine learning that measures the ability of a model to identify all relevant instances within a dataset. In the context of object detection or classification, it specifically answers the question: "Out of all the actual positive cases, how many did the model correctly find?" Achieving high recall is critical in scenarios where missing a positive instance—often referred to as a false negative—carries significant consequences. Unlike accuracy, which can be misleading when dealing with imbalanced data, recall provides a focused view on the model's effectiveness at "capturing" the target class.

L'importanza di un'alta recall

In many artificial intelligence applications, the cost of failing to detect an object is far higher than the cost of a false alarm. A model optimized for recall minimizes false negatives, ensuring that the system casts a wide enough net to catch potential threats, anomalies, or critical conditions. This often involves a trade-off, as increasing recall can sometimes lead to a lower precision score, meaning the model might flag more non-relevant items as positive. Understanding this balance is key to developing robust machine learning solutions.

Applicazioni nel mondo reale

Il richiamo è la metrica trainante alla base di molte soluzioni di intelligenza artificiale critiche per la sicurezza . Ecco due esempi significativi in cui la sensibilità ha la precedenza:

• Diagnostica detect : nell' analisi delle immagini mediche, come lo screening radiografico per individuare i primi segni di malattia, un'elevata sensibilità è imprescindibile. Se un sistema di IA nel settore sanitario viene utilizzato per individuare tumori, è molto meglio che il sistema segnali un'ombra sospetta che si rivela benigna (un falso positivo) piuttosto che non individuare affatto un tumore maligno. I medici si affidano a questi strumenti come rete di sicurezza, per garantire che nessun potenziale rischio per la salute venga trascurato.
• Sicurezza e sorveglianza: per un sistema di allarme di sicurezza, l'obiettivo principale è detect tentativo di intrusione. Un sistema ottimizzato per un elevato richiamo garantisce che, se una persona entra in una zona riservata , l'allarme si attivi. Anche se questo potrebbe portare a occasionali falsi allarmi causati dalla fauna selvatica, è preferibile che il sistema non riesca a detect intruso reale. I modelli di rilevamento degli oggetti in questi scenari sono ottimizzati per garantire la massima sensibilità alle potenziali minacce.

Recall vs. Precision

It is essential to distinguish recall from its counterpart, precision. While recall measures the quantity of relevant cases found (completeness), precision measures the quality of the positive predictions (exactness).

• Recall: Focuses on avoiding missed detections. "Did we find all the apples?"
• Precision: Focuses on minimizing false alarms. "Are all the things we called apples actually apples?"

These two metrics often share an inverse relationship, visualized through a Precision-Recall curve. To evaluate the overall balance between them, developers often look at the F1-score, which is the harmonic mean of both. In imbalanced datasets, looking at recall alongside the confusion matrix gives a much clearer picture of performance than accuracy alone.

Misurare il richiamo con Ultralytics YOLO

When training models like the cutting-edge YOLO26, recall is automatically computed during the validation phase. The framework calculates recall for each class and the mean Average Precision (mAP), helping developers gauge how well the model finds objects.

È possibile convalidare facilmente un modello addestrato e visualizzarne le metriche di richiamo utilizzando Python. Questo snippet mostra come caricare un modello e verificarne le prestazioni su un set di dati standard:

from ultralytics import YOLO

# Load a pretrained YOLO26 model
model = YOLO("yolo26n.pt")

# Validate the model on the COCO8 dataset
# The results object contains metrics like Precision, Recall, and mAP
metrics = model.val(data="coco8.yaml")

# Access and print the mean recall score for box detection
print(f"Mean Recall: {metrics.results_dict['metrics/recall(B)']:.4f}")

This code utilizes the Ultralytics API to run validation. If the recall is lower than required for your project, you might consider techniques like data augmentation to create more varied training examples or hyperparameter tuning to adjust the model's sensitivity. Using the Ultralytics Platform can also streamline the process of managing datasets and tracking these metrics over multiple training runs.

Improving Model Recall

To boost a model's recall, data scientists often adjust the confidence threshold used during inference. Lowering the threshold makes the model more "optimistic," accepting more predictions as positive, which increases recall but may decrease precision. Additionally, collecting more diverse training data helps the model learn to recognize hard negatives and obscure instances. For complex tasks, employing advanced architectures like Transformer blocks or exploring ensemble methods can also improve the system's ability to detect subtle features that simpler models might miss.