.. _new: What's New in FindFace CIBR 2.1.1 =================================== .. rubric:: New Features: * :ref:`Upload a photo ` to a case: the end user is now able to upload only photos, videos, or photos along with videos to a case. * Rematching clusters with the record index: once new photos are uploaded to the record index, the system checks them against all existing clusters of open cases to find out new matches. * "Inbox" of cases: the system marks cases that have at least one unacknowledged case cluster with a special sign so that the end user does not miss it. See :ref:`case-files` .. rubric:: Technical Changes: * Technical architecture: microservices instead of monolith. See :ref:`architecture` * :ref:`Docker-based ` distribution: FindFace CIBR 2.1.1 runs as a set of docker containers described in a ``docker-compose`` file. * Deployment: FindFace CIBR is distributed with its own :ref:`installer ` now. * CentOS / Debian support. .. rubric:: Enhanced Algorithms, UI, UX: * Improvements to the mechanism of clustering and more flexible work with clusters: clusters can now be manually merged and unmerged. * Record index :ref:`batch upload ` reinstated. * Enhanced :ref:`Daily search ` functionality: last 24-hour' location registration of the objects of interest in remote monitoring. .. rubric:: Fresh Neural Networks: * Face detection: a new face detection model ``face.jasmine_fast.003`` with improved characteristics. * Face image normalization: new face image normalization models ``facenorm.multicrop_full_center_size400`` and ``facenorm.multicrop_full_crop2x_size400`` that use smarter normalization algorithm and a new version ``bee.v3`` with improved characteristics. * Face recognition: brand-new ``nectarine_xs_320`` and ``nectarine_m_160`` face recognition models, much faster and more accurate than preceding models. * Face quality recognition: a new version ``quality_fast.v1`` that predicts face quality to get the best shot and filter trash.