Object classification is a standout: people, animals, vehicles, and packages are usually identified accurately, which helps reduce noisy motion alerts.
Object classification is a standout: reviewers repeatedly report accurate person and package identification, plus useful filtering for animals and vehicles, helping keep alerts relevant.
Person/human detection is repeatedly emphasized as effective and valuable, especially because reviewers contrast it with competitors that require subscriptions for similar person-detection features.
Person and object detection is generally effective for visitors and delivered boxes, but classification can be imperfect with unusual objects or when package detection is too sensitive.
Person/package identification is a major reason reviewers recommend the E340, with consistent mentions of human and package detection. Several sources still note occasional misreads, but the consensus is that detection is useful and generally effective for real-world deliveries.
Person, pet, and vehicle recognition is frequently highlighted as working without a subscription, improving alert relevance; package-specific detection is consistently noted as missing on the D210.
Person detection is generally reported as accurate and useful, and is often cited as a core strength. Package detection quality varies more than person detection.
Object recognition via HomeKit Secure Video can distinguish people, animals, and vehicles; accuracy is generally praised but not perfect, especially in tricky nighttime scenes.
Smart recognition (people, packages, animals, vehicles) is viewed as helpful and often accurate when enabled, but it is usually locked behind the subscription.
With HomeKit Secure Video and/or Aqara alerts, the doorbell can distinguish people and other motion types, but real-world consistency varies and can depend on lighting, placement, and whether you are using HSV or Aqara-native recording.
Person and object detection can add context to alerts. Multiple reviews mention people detection and some also note vehicle detection in event logs or notifications.
Person/human detection is consistently discussed as a core behavior, while broader categories (pet/vehicle) show up as available in certain configurations (notably with HomeBase). One reviewer notes package detection is not part of the system they tested.
People-only or person alerts are effective for cutting noise, but reviewers also note it is not as information-rich as top-tier competitors and some wish for dedicated package detection.
Smart detections (people and other object categories) are repeatedly referenced as available, but typically tied to the Roku Smart Home subscription rather than the free tier.
Person and package detection is frequently presented as accurate and useful, especially for deliveries. Multiple sources note these smarter alerts are often tied to a subscription tier rather than fully free.
Person detection is repeatedly referenced as helpful for reducing false alerts. Package and vehicle detection are discussed as available in some setups or versions, but reviewers also describe gaps and inconsistency depending on model and software.
Detection by type is useful when it works, but package labeling accuracy is described as inconsistent by some, especially compared with premium rivals.
Smart detection can label people, packages, pets, and vehicles with Cam Plus; accuracy is generally best for people and less consistent for vehicles/packages.
Detection is largely centered on people (with some references to car filtering), and reviewers repeatedly call out missing package/animal/vehicle detection and face recognition compared to newer premium doorbells.
Person detection is available, but frequently described as subscription-gated and not always perfect; some sources also mention vehicle filtering. Package detection and facial recognition are generally absent.
Detection is basic and does not reliably distinguish people, packages, animals, or cars. Reviews repeatedly note the absence of person or package intelligence compared with pricier competitors.