PETR: Rethinking the Capability of Transformer-Based Language Model in Scene Text Recognition.

The exploration of linguistic information promotes the development of scene text recognition task. Benefiting from the significance in parallel reasoning and global relationship capture, transformer-based language model (TLM) has achieved dominant performance recently. As a decoupled structure from the recognition process, we argue that TLM's capability is limited by the input low-quality visual prediction. To be specific: 1) The visual prediction with low character-wise accuracy increases the correction burden of TLM. 2) The inconsistent word length between visual prediction and original image provides a wrong language modeling guidance in TLM. In this paper, we propose a Progressive scEne Text Recognizer (PETR) to improve the capability of transformer-based language model by handling above two problems. Firstly, a Destruction Learning Module (DLM) is proposed to consider the linguistic information in the visual context. DLM introduces the recognition of destructed images with disordered patches in the training stage. Through guiding the vision model to restore patch orders and make word-level prediction on the destructed images, visual prediction with high character-wise accuracy is obtained by exploring inner relationship between the local visual patches. Secondly, a new Language Rectification Module (LRM) is proposed to optimize the word length for language guidance rectification. Through progressively implementing LRM in different language modeling steps, a novel progressive rectification network is constructed to handle some extremely challenging cases (e.g. distortion, occlusion, etc.). By utilizing DLM and LRM, PETR enhances the capability of transformer-based language model from a more general aspect, that is, focusing on the reduction of correction burden and rectification of language modeling guidance. Compared with parallel transformer-based methods, PETR obtains 1.0% and 0.8% improvement on regular and irregular datasets respectively while introducing only 1.7M additional parameters. The extensive experiments on both English and Chinese benchmarks demonstrate that PETR achieves the state-of-the-art results.

Automated counting of bacterial colonies on agar plates based on images captured at near-infrared light.

Counting colonies is usually used in microbiological analysis to assess if samples meet microbiological criteria. Although manual counting remains gold standard, the process is subjective, tedious, and time-consuming. Some developed automatic counting methods could save labors and time, but their results are easily affected by uneven illumination and reflection of visible light. To offer a method which counts colonies automatically and is robust to light, we constructed a convenient and cost-effective system to obtain images of colonies at near-infrared light, and proposed an automatic method to detect and count colonies by processing images. The colonies cultured by using raw cows' milk were used as identification objects. The developed system mainly consisted of a visible/near-infrared camera and a circular near-infrared illuminator. The automatic method proposed to count colonies includes four steps, i.e., eliminating noises outside agar plate, removing plate rim and wall, identifying and separating clustered or overlapped colonies, and counting colonies by using connected region labelling, distance transform, and watershed algorithms, etc. A user-friendly graphic user interface was also developed for the proposed method. The relative error and counting time of the automatic counting method were compared with those of manual counting. The results showed that the relative error of the automatic counting method was -7.4%~ + 8.3%, with average relative error of 0.2%, and the time used for counting colonies on each agar plate was 11-21 s, which was 15-75% of the time used in manual counting, depending on the numbers of colonies on agar plates. The proposed system and automatic counting method demonstrate promising performance in terms of precision, and they are robust and efficient in terms of labor- and time-savings.