Character-based Identification Research Articles

An adaptive n-gram transformer for multi-scale scene text recognition

While vision transformers have been highly successful in improving the performance in image-based tasks, not much work has been reported on applying transformers to scene text recognition due to the complexities in the visual appearance of multi-scale texts. To fill the gap, this paper proposes an adaptive n-gram transformer for multi-scale scene text recognition (ANT-STR). In ANT-STR, an adaptive n-gram embedding that is able to automatically determine the optimal size of each image patch is designed to fully explore the potential semantic correlations between neighboring visual patches, which is essential for feature extraction from multi-scale scene texts. On top of the adaptive n-gram embedding, a patch-based n-gram attention mechanism is introduced into ANT-STR to further process the feature maps for multi-scale texts. In addition, the loss function is rectified to take into account both multi-scale character-based identification and contextual coherence scoring. Comparative studies are conducted on five widely used benchmark datasets and a new multi-scale scene text dataset collected from tourism scenes in Indonesia. Our experimental results demonstrate that ANT-STR performs considerably better compared to the state-of-the-art, especially in handling complex multi-scale scene texts.

Species-specific molecular signatures for the commercially important scombrids using mitochondrial gene analysis; a tool for fisheries management

The family Scombridae, comprising of commercially important marine fishes, which are widely distributed across tropical, subtropical, and temperate waters. Scombrids constitute an important fishery and are exploited across their range of distribution. Accurate identification of species is important for effective fisheries management and forensics purposes. In the present study, commercially important species of scombrids distributed in Indian waters were selected for molecular phylogeny and development of species-specific diagnostic keys using mitochondrial cytochrome c oxidase I (COI) gene. In addition to morphological identification, the species were identified using molecular-based identification approach and character-based identification system. Molecular phylogeny for 16 species of scombrids was ascertained using mitochondrial COI analyses. The character-based molecular diagnostic keys for 16 species of scombrids were developed in-silico from data set comprising 603 mtCOI sequences generated in the present study as well as collected from published resources. The study uncovered 39 unambiguous molecular keys, capable of distinguishing all reference taxa under study. The development of species-specific sequences and key diagnostic nucleotides for scombrid fish species will be helpful in effective forensic applications, enforcement of fisheries regulations, and for sustainable utilization of this commercially important group of species.

Character-based diagnostic keys, molecular identification and phylogenetic relationships of threadfin breams (family: Nemipteridae) based on mitochondrial genes from the Southern coromandel Coast, India

DNA barcoding, primarily focusing on cytochrome c oxidase subunit I (COI) gene has been appraised as an effective tool for species identification. In this study, we focused on the marine fishes of Family Nemipteridae, one of the commercially important group distributed within the Coromandel Coast. The Partial sequences of COI and 16S rRNA of mitochondrial genes were analyzed for species identification and phylogenetic relationship of Nemipterus species (Nemipterus japonicus, Nemipterus peronii, Nemipterus bipunctatus, Nemipterus bathybius). Character-based identification approaches that categorize specimens to species using classification rules that compactly identify species in terms of key diagnostic nucleotides in selected gene sequences. Using the BLOG 2.0 software, species-specific diagnostic nucleotides were identified for the selected group of species. A data set of 198 mtCOI sequences was obtained from published resources and used to screen character-based molecular diagnostic keys for species in silico. Partial sequences of both the genes provided sufficient phylogenetic information to distinguish the four Nemipterus species indicating the usefulness of mtDNA-based approach in species identification. This study proves the use of mtDNA genes sequence-based approach is a support tool along with traditional taxonomy for identifying fish species at a faster pace.

Character-based identification key for commercially important fishes of Pulicat lake: tool for conservation and management

Fishes are an important group of vertebrates in the animal world and make a significant contribution to global biodiversity. Fish is used as a source of food and contains many essential vitamins and fatty acids. The study of fish and their stability is important because, from year to year, fish stocks are often very important. For the conservation and management of these dwindling resources, correct identification of species is a prerequisite. Character-based methods of identification are of considerable use in this context, which classify specimens into species using classification rules that compactly describe species in terms of key diagnostic nucleotides in the gene sequences chosen. In this study, a total of 56 species of fishes distributed in Pulicat lake waters is taken as the target group. Mitochondrial CO1 sequences of each species were downloaded and modified. The species-specific diagnostic nucleotides for the selected group of species were identified using the BLOG version 2.0 software. Species-specific probes with a length range of 18–37 bp were designed on the basis of identified diagnostic nucleotide sites. The method is an effective tool for quickly and easily obtaining a significant amount of reliable information and could be used for forensic applications and conservation of fishes in Pulicat Lake.

Character-based identification system of scombrids from Indian waters for authentication and conservation purposes

Scombrids are the important component of pelagic fishery resources which include 54 species under 15 genera commonly known as mackerels, bonitos, and tunas. Due to the high commercial value attained, there are real chances of fraudulent substitution by species of inferior value. DNA based species identification methods can be applied to detect product adulteration, as well as to better contribute to the conservation and management of these species by providing accurate species identification independently of the age of the individuals or the tissue processed. In this study, a total of 15 commercially important scombrid species from Indian waters were analyzed. Due to the inadequacy of mitochondrial COI barcoding gene in discriminating between some Thunnus species, cytochrome b sequences were used instead. For all the 15 species, we propose a DNA character-based keys which uses a diagnostic combination of nucleotides and respective probes, including the first character-based keys and probes to differentiate between Thunnus albacares and T. obsesus.

Character based identification system for Elasmobranchs for conservation and forensic applications

The term Elasmobranchs or Chondrichthyans refer to the group of marine organisms include sharks, skates, rays and chimeras. Due to high demand most of the Elasmobranch species depleted drastically in recent years. Being the apex predators in the marine environment most of the Chondrichthyan species are vulnerable with some risk of extinction. Accurate identification of species is prerequisite for conservation and management of these declining resources. In this context, character-based identification methods are of immense use which classify specimen to species using classification rules that compactly characterized species in terms of key diagnostic nucleotides in the selected gene sequences. In the present study, a total of 82 species of elasmobranches distributed in Indian waters is taken as the target group. A total of 20–30 mitochondrial CO1 sequences of each species were downloaded and modified. Using the BLOG2.0 software, species specific diagnostic nucleotides were identified for the selected group of species. Based on identified diagnostic nucleotide sites, species-specific probes with a length range of 23–30 bp were designed. These probes could be useful in detecting species of interest in customized microarrays (DNA chip). The technique is a powerful tool to obtain a significant amount of accurate information quickly and conveniently and could be used for forensic applications and conservation of elasmobranches in Indian waters.

Effective language identification of forum texts based on statistical approaches

This investigation deals with the problem of language identification of noisy texts, which could represent the primary step of many natural language processing or information retrieval tasks. Language identification is the task of automatically identifying the language of a given text. Although there exists several methods in the literature, their performances are not so convincing in practice.In this contribution, we propose two statistical approaches: the high frequency approach and the nearest prototype approach. In the first one, 5 algorithms of language identification are proposed and implemented, namely: character based identification (CBA), word based identification (WBA), special characters based identification (SCA), sequential hybrid algorithm (HA1) and parallel hybrid algorithm (HA2). In the second one, we use 11 similarity measures combined with several types of character N-Grams.For the evaluation task, the proposed methods are tested on forum datasets containing 32 different languages. Furthermore, an experimental comparison is made between the proposed approaches and some referential language identification tools such as: LIGA, NTC, Google translate and Microsoft Word. Results show that the proposed approaches are interesting and outperform the baseline methods of language identification on forum texts.

Molecular Identification of Atlantic Bluefin Tuna (Thunnus thynnus, Scombridae) Larvae and Development of a DNA Character-Based Identification Key for Mediterranean Scombrids.

The Atlantic bluefin tuna, Thunnus thynnus, is a commercially important species that has been severely over-exploited in the recent past. Although the eastern Atlantic and Mediterranean stock is now showing signs of recovery, its current status remains very uncertain and as a consequence their recovery is dependent upon severe management informed by rigorous scientific research. Monitoring of early life history stages can inform decision makers about the health of the species based upon recruitment and survival rates. Misidentification of fish larvae and eggs can lead to inaccurate estimates of stock biomass and productivity which can trigger demands for increased quotas and unsound management conclusions. Herein we used a molecular approach employing mitochondrial and nuclear genes (CO1 and ITS1, respectively) to identify larvae (n = 188) collected from three spawning areas in the Mediterranean Sea by different institutions working with a regional fisheries management organization. Several techniques were used to analyze the genetic sequences (sequence alignments using search algorithms, neighbour joining trees, and a genetic character-based identification key) and an extensive comparison of the results is presented. During this process various inaccuracies in related publications and online databases were uncovered. Our results reveal important differences in the accuracy of the taxonomic identifications carried out by different ichthyoplanktologists following morphology-based methods. While less than half of larvae provided were bluefin tuna, other dominant taxa were bullet tuna (Auxis rochei), albacore (Thunnus alalunga) and little tunny (Euthynnus alletteratus). We advocate an expansion of expertise for a new generation of morphology-based taxonomists, increased dialogue between morphology-based and molecular taxonomists and increased scrutiny of public sequence databases.

What barcode sequencing reveals about the shark fishery in Peru

Elasmobranchs are rapidly declining due to overfishing and bycatch, underlining the need for immediate protection. Critical baseline information on the diversity of targeted species is, however, often missing. Peru is a major country for shark fishery, an activity that has been under-regulated and poorly monitored, aggravated by the superficial taxonomic identifications at landing points across the country. Furthermore, most of the species landed by the shark fishery in Peru are listed as Vulnerable in the IUCN Red List. To assess the diversity of shark species targeted by fisheries in Peru we analyzed the variation of the cytochrome oxidase I (cox1) region of the mitochondrial DNA from 118 samples collected between 2004 and 2009, from six landing points. Our analysis revealed unambiguously that the 16 shark species classified by fishermen using meristic characters corresponded only to nine species. While some commonly landed species (e.g. Prionace glauca) were consistently correctly identified, for others species multiple inconsistent names were applied (e.g. Galeorhinus galeus). Our molecular characterization further allowed the identification of specimens with non-informative common names (i.e. “tiburon”=shark). In most cases the unknown specimens were Isurus oxyrinchus and P. glauca. Interestingly, all samples labeled as common thresher (Alopias vulpinus) were identified as pelagic thresher (Alopias pelagicus). Finally, one sample was equivocally identified as a dusky shark (Carcharhinus obscurus) and as a galapagos shark (Carcharhinus galapagensis) reinforcing the genetic similarity reported for these species. We generated a character-based identification library containing 26 of the 31 commercially important sharks landed in Peru and tested its performance as a species diagnostic. The library correctly identified 25 out of 28 barcodes tested, outperforming the distance-based approach. This is the first study sequencing barcodes of marine species in Peru and generated a genetic reference library of targeted shark species. We suggest that the molecular tools used are a quick and effective complement for the monitoring of the fishery of threatened shark species. A combined effort to obtain these data, by countries in the east Pacific region with an on-going shark fishery, would provide with the essential guiding information to promote the implementation of effective sustainable management plans.

Authentication of raw and processed tuna from Indonesian markets using DNA barcoding, nuclear gene and character-based approach

Establishing seafood authentication methods is an important task for fisheries research laboratories and food control authorities. Nowadays, the extent of fish species substitution is suspected being greater than ever before in commercial markets. In order to provide reliable polymerase chain reaction (PCR)-based authentication systems for tunas, we collected and analyzed authentic tuna reference samples and tuna-food products from Indonesian markets. Our analytical methods mainly relied on identification using the mitochondrial cytochrome c oxidase subunit I (COI) gene, as a genetic marker for “DNA barcoding,” as well as the rhodopsin (RH1) gene as a nuclear marker. Additionally, we identified species-specific nucleotide diagnostic positions (characters) to complete the results obtained basic local alignment search (BLAST) and phylogenetic analysis. Authentication results of tuna-food products showed relatively successful amplification for the COI gene; RH1 acted as an alternative solution for some of the samples, which had failed to react in COI-PCR. Species of the genus Thunnus could not be unambiguously differentiated by BLAST and phylogenetic analysis (neighbor-joining tree) in all cases due to the high similarity of the COI sequences. However, the character-based identification method was found to be helpful for species assignment in case of tuna-food products. Therefore, our findings demonstrated that the COI gene could be more reliable used as a tool for Indonesian commercial tuna products authentication, if the sequencing results were combined with the character-based identification using differences at certain nucleotide positions.

Multi-gene barcoding to discriminate sibling species within a morphologically difficult fish genus (Sillago)

Fisheries species that cannot be reliably identified based on unique morphological features pose a challenge for research, monitoring and management. DNA barcoding, which refers to a standardized procedure for genetic species identification based on the mitochondrial cytochrome oxidase 1 (CO1) gene, is an important tool to confront this problem. Here, we present a multi-gene barcoding approach to discriminate two fisheries species flagged as sibling species within the morphologically difficult fish genus Sillago: S analis and S. ciliata. CO1 revealed low interspecific variation that was insufficient for genetic distance-based species identification. Despite this, one diagnostic site did permit genetic character-based identification. However, four cases of mismatch among morphological identifications (mIDs) and CO1 barcodes required analysing additional genes and loci. A spot test approach was used to screen five other genes in the mitochondrial genome (mtDNA) and four genes in the nuclear genome (nuDNA). The nuDNA recombination activating gene 2 (RAG2) revealed most diagnostic sites, correctly clustering mIDs of all individuals using both genetic distance- and character-based analysis. Three of four cases of mID/CO1 mismatches were found to be simultaneous CO1/RAG2 (mtDNA/nuDNA) mismatches indicative of past hybridization. The fourth mID/CO1 mismatch was found to represent an initial morphological misidentification. Five spot-tested genes, containing either fully or partially diagnostic sites (mtDNA: NADH dehydrogenase 2 [ND2], ATP synthase [ATPase] and cytochrome b; nuDNA: histone protein 3a and ring finger protein 213), confirmed these findings. Based on our preliminary data, we recommend (1) that S. analis and S. ciliata retain their taxonomic status as sibling species with full corresponding recognition as separate management and conservation units, and (2) precautionary management of S. ciliata fisheries until further research into hybridization is carried out. The multi-gene spot test approach and efficient markers (RAG2 in combination with ND2, ATPase or CO1) may help identifying other problematic fish species.

Anopheline species composition in borderline of Iran–Azerbaijan

Malaria is still one of the most important health-problems in the world and is endemic in Iran. Since 1994, after collapse of former Soviet Union, a new threat of malaria importation emerged from those countries into the northern Iran. This work was carried out to provide further evidence on the status of anopheline species composition, the malaria parasite species, and natural infectivity of mosquitoes distributed in Pars-Abad district, on the borderline of Azerbaijan in northwestern Iran. Mosquitoes were collected from May to December 2008 in anopheline seasonal activity and were identified at the species level. The genus- and species-specific primers against Plasmodium ssrDNA gene were used for specific amplification on female mosquito head+thorax. Members of the Anoheles maculipennis complex were identified by sequence analysis of the ribosomal internal transcribed spacer II (ITS2-rDNA). Morphological character-based identification showed that out of 1455 anopheline female specimens, 1121 (77%) were of A. maculipennis s.l. and 334 (23%) were of Anoheles hyrcanus. Molecular analysis of the species complex indicated the presence of Anoheles sacharovi 984 (67.6%) and A. maculipennis 137 (9.4%) in the region. None of themosquito's head–thorax was found to be naturally infected by malaria parasite. Results of this study, particularly high dominance of A. sacharovi, suggest a potential risk of malaria epidemic in the region, and the need for a continuous epidemiological surveillance.

The development and evaluation of a single step multiplex PCR for simultaneous detection of Anopheles annularis group mosquitoes, human host preference and Plasmodium falciparum sporozoite presence

The Anopheles annularis group mosquitoes, subgenus Cellia Theobald (Diptera: Culicidae), includes five recognized species: An. annularis Van der Wulp, An. nivipes Theobald, An. pallidus Theobald, An. philippinensis Ludlow and An. schueffneri Stanton. From these five, the three most common species found in Orissa were considered for this study because of their remarkable vectorial and behavioral variation and the important role they play in malaria transmission. To identify and understand their role in malaria transmission we developed a single multiplex PCR-based assay. This assay included the detection of human blood feeding habit and Plasmodium falciparum sporozoite presence. Of the 186 An. annularis mosquitoes collected, morphological character-based identification showed that 94 were An. annularis, 54 were An. philippinensis and 38 were An. pallidus. However, the multiplex PCR assay confirmed that 91 were An. annularis, 56 were An. philippinensis and 39 were An. pallidus individuals after adjustments were made for misidentified specimens in the morphological method. Anopheles annularis and An. philippinensis were found positive for human blood, and two samples of An. annularis species were positive for P. falciparum sporozoites. This one-step PCR-based method constitutes a very powerful tool in large surveys of anopheline populations.