A Systematic Investigation on end-to-end Deep Recognition of Grocery Products in the Wild

2021 ◽  
Author(s):  
Marco Leo ◽  
Pierluigi Carcagni ◽  
Cosimo Distante
Keyword(s):  
Systematic Investigation ◽  
In The Wild ◽  
End To End ◽  
Grocery Products

scholarly journals Blocks in the pseudouridimycin pathway unlock hidden metabolites in the Streptomyces producer strain

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Marianna Iorio ◽  
Sahar Davatgarbenam ◽  
Stefania Serina ◽  
Paolo Criscenzo ◽  
Mitja M. Zdouc ◽  
...  
Keyword(s):  
Gene Clusters ◽  
Biosynthetic Gene Cluster ◽  
Systematic Investigation ◽  
Biosynthetic Gene ◽  
Wild Type ◽  
Bacterial Rna ◽  
Soil Isolate ◽  
Mutant Strains ◽  
In The Wild ◽  
Polymerase Inhibitor

AbstractWe report a metabolomic analysis of Streptomyces sp. ID38640, a soil isolate that produces the bacterial RNA polymerase inhibitor pseudouridimycin. The analysis was performed on the wild type, on three newly constructed and seven previously reported mutant strains disabled in different genes required for pseudouridimycin biosynthesis. The results indicate that Streptomyces sp. ID38640 is able to produce, in addition to lydicamycins and deferroxiamines, as previously reported, also the lassopeptide ulleungdin, the non-ribosomal peptide antipain and the osmoprotectant ectoine. The corresponding biosynthetic gene clusters were readily identified in the strain genome. We also detected the known compound pyridindolol, for which we propose a previously unreported biosynthetic gene cluster, as well as three families of unknown metabolites. Remarkably, the levels of most metabolites varied strongly in the different mutant strains, an observation that enabled detection of metabolites unnoticed in the wild type. Systematic investigation of the accumulated metabolites in the ten different pum mutants identified shed further light on pseudouridimycin biosynthesis. We also show that several Streptomyces strains, able to produce pseudouridimycin, have distinct genetic relationship and metabolic profile with ID38640.

scholarly journals Frequency-domain loss function for deep exposure correction of dark images

2021 ◽  
Author(s):  
Ojasvi Yadav ◽  
Koustav Ghosal ◽  
Sebastian Lutz ◽  
Aljosa Smolic
Keyword(s):  
Loss Function ◽  
Correct Choice ◽  
Frequency Space ◽  
Multi Scale ◽  
Quantitative Metrics ◽  
Image Translation ◽  
Domain Loss ◽  
In The Wild ◽  
End To End ◽  
Classical Image

AbstractWe address the problem of exposure correction of dark, blurry and noisy images captured in low-light conditions in the wild. Classical image-denoising filters work well in the frequency space but are constrained by several factors such as the correct choice of thresholds and frequency estimates. On the other hand, traditional deep networks are trained end to end in the RGB space by formulating this task as an image translation problem. However, that is done without any explicit constraints on the inherent noise of the dark images and thus produces noisy and blurry outputs. To this end, we propose a DCT/FFT-based multi-scale loss function, which when combined with traditional losses, trains a network to translate the important features for visually pleasing output. Our loss function is end to end differentiable, scale-agnostic and generic; i.e., it can be applied to both RAW and JPEG images in most existing frameworks without additional overhead. Using this loss function, we report significant improvements over the state of the art using quantitative metrics and subjective tests.

scholarly journals A Data Driven End-to-End Approach for In-the-Wild Monitoring of Eating Behavior Using Smartwatches

2021 ◽  
Vol 25 (1) ◽  
pp. 22-34 ◽  
Author(s):  
Konstantinos Kyritsis ◽  
Christos Diou ◽  
Anastasios Delopoulos
Keyword(s):  
Eating Behavior ◽  
Data Driven ◽  
In The Wild ◽  
End To End

Deep tracking in the wild: End-to-end tracking using recurrent neural networks

2017 ◽  
Vol 37 (4-5) ◽  
pp. 492-512 ◽  
Author(s):  
Julie Dequaire ◽  
Peter Ondrúška ◽  
Dushyant Rao ◽  
Dominic Wang ◽  
Ingmar Posner
Keyword(s):  
Object Tracking ◽  
Autonomous Vehicle ◽  
Urban Environments ◽  
Positive Contribution ◽  
Efficient Manner ◽  
Model Free ◽  
Novel Approach ◽  
In The Wild ◽  
End To End ◽  
Traditional Approaches

This paper presents a novel approach for tracking static and dynamic objects for an autonomous vehicle operating in complex urban environments. Whereas traditional approaches for tracking often feature numerous hand-engineered stages, this method is learned end-to-end and can directly predict a fully unoccluded occupancy grid from raw laser input. We employ a recurrent neural network to capture the state and evolution of the environment, and train the model in an entirely unsupervised manner. In doing so, our use case compares to model-free, multi-object tracking although we do not explicitly perform the underlying data-association process. Further, we demonstrate that the underlying representation learned for the tracking task can be leveraged via inductive transfer to train an object detector in a data efficient manner. We motivate a number of architectural features and show the positive contribution of dilated convolutions, dynamic and static memory units to the task of tracking and classifying complex dynamic scenes through full occlusion. Our experimental results illustrate the ability of the model to track cars, buses, pedestrians, and cyclists from both moving and stationary platforms. Further, we compare and contrast the approach with a more traditional model-free multi-object tracking pipeline, demonstrating that it can more accurately predict future states of objects from current inputs.

scholarly journals EnsNet: Ensconce Text in the Wild

2019 ◽  
Vol 33 ◽  
pp. 801-808 ◽  
Author(s):  
Shuaitao Zhang ◽  
Yuliang Liu ◽  
Lianwen Jin ◽  
Yaoxiong Huang ◽  
Songxuan Lai
Keyword(s):  
Generative Adversarial Network ◽  
Local Consistency ◽  
Adversarial Network ◽  
Image Patches ◽  
Scene Text ◽  
In The Wild ◽  
Lateral Connection ◽  
Previous State ◽  
End To End ◽  
General Object

A new method is proposed for removing text from natural images. The challenge is to first accurately localize text on the stroke-level and then replace it with a visually plausible background. Unlike previous methods that require image patches to erase scene text, our method, namely ensconce network (EnsNet), can operate end-to-end on a single image without any prior knowledge. The overall structure is an end-to-end trainable FCN-ResNet-18 network with a conditional generative adversarial network (cGAN). The feature of the former is first enhanced by a novel lateral connection structure and then refined by four carefully designed losses: multiscale regression loss and content loss, which capture the global discrepancy of different level features; texture loss and total variation loss, which primarily target filling the text region and preserving the reality of the background. The latter is a novel local-sensitive GAN, which attentively assesses the local consistency of the text erased regions. Both qualitative and quantitative sensitivity experiments on synthetic images and the ICDAR 2013 dataset demonstrate that each component of the EnsNet is essential to achieve a good performance. Moreover, our EnsNet can significantly outperform previous state-of-the-art methods in terms of all metrics. In addition, a qualitative experiment conducted on the SBMNet dataset further demonstrates that the proposed method can also preform well on general object (such as pedestrians) removal tasks. EnsNet is extremely fast, which can preform at 333 fps on an i5-8600 CPU device.

scholarly journals Structure and Distribution of Endogenous Nonecotropic Murine Leukemia Viruses in Wild Mice

1998 ◽  
Vol 72 (10) ◽  
pp. 8289-8300 ◽  
Author(s):  
Keizo Tomonaga ◽  
John M. Coffin
Keyword(s):  
Inbred Mice ◽  
Evolutionary Relationship ◽  
Wild Mouse ◽  
Systematic Investigation ◽  
Endogenous Retroviruses ◽  
Wild Mice ◽  
In The Wild ◽  
Leukemia Viruses ◽  
Specific Sequences ◽  
Murine Leukemia

ABSTRACT Virtually all of our present understanding of endogenous murine leukemia viruses (MLVs) is based on studies with inbred mice. To develop a better understanding of the interaction between endogenous retroviruses and their hosts, we have carried out a systematic investigation of endogenous nonecotropic MLVs in wild mice. Species studied included four major subspecies of Mus musculus(M. m. castaneus, M. m. musculus, M. m. molossinus, and M. m. domesticus) as well as four common inbred laboratory strains (AKR/J, HRS/J, C3H/HeJ, and C57BL/6J). We determined the detailed distribution of nonecotropic proviruses in the mice by using both env- and long terminal repeat (LTR)-derived oligonucleotide probes specific for the three different groups of endogenous MLVs. The analysis indicated that proviruses that react with all of the specific probes are present in most wild mouse DNAs tested, in numbers varying from 1 or 2 to more than 50. Although in common inbred laboratory strains the linkage of group-specific sequences in env and the LTR of the proviruses is strict, proviruses which combine env and the LTR sequences from different groups were commonly observed in the wild-mouse subspecies. The “recombinant” nonecotropic proviruses in the mouse genomes were amplified by PCR, and their genetic and recombinant natures were determined. These proviruses showed extended genetic variation and provide a valuable probe for study of the evolutionary relationship between MLVs and the murine hosts.

scholarly journals Scale-Invariant Multidirectional License Plate Detection with the Network Combining Indirect and Direct Branches

Sensors ◽  
2021 ◽  
Vol 21 (4) ◽  
pp. 1074
Author(s):  
Song-Lu Chen ◽  
Qi Liu ◽  
Jia-Wei Ma ◽  
Chun Yang
Keyword(s):  
Direct Detection ◽  
Direct Methods ◽  
Input Image ◽  
Indirect Detection ◽  
License Plate ◽  
Scale Invariant ◽  
Refinement Method ◽  
In The Wild ◽  
License Plate Detection ◽  
End To End

As the license plate is multiscale and multidirectional in the natural scene image, its detection is challenging in many applications. In this work, a novel network that combines indirect and direct branches is proposed for license plate detection in the wild. The indirect detection branch performs small-sized vehicle plate detection with high precision in a coarse-to-fine scheme using vehicle–plate relationships. The direct detection branch detects the license plate directly in the input image, reducing false negatives in the indirect detection branch due to the miss of vehicles’ detection. We propose a universal multidirectional license plate refinement method by localizing the four corners of the license plate. Finally, we construct an end-to-end trainable network for license plate detection by combining these two branches via post-processing operations. The network can effectively detect the small-sized license plate and localize the multidirectional license plate in real applications. To our knowledge, the proposed method is the first one that combines indirect and direct methods into an end-to-end network for license plate detection. Extensive experiments verify that our method outperforms the indirect methods and direct methods significantly.

scholarly journals Scent marking in Sunda clouded leopards (Neofelis diardi): Novel observations close a key gap in understanding felid communication behaviours

2020 ◽  
Author(s):  
ML Allen ◽  
Heiko Wittmer ◽  
E Setiawan ◽  
S Jaffe ◽  
AJ Marshall
Keyword(s):  
Scent Marking ◽  
Systematic Investigation ◽  
Behavioural Ecology ◽  
Natural Environments ◽  
Basic Information ◽  
Video Cameras ◽  
In The Wild ◽  
Intraspecific Communication ◽  
Communication Behaviours ◽  
Felid Species

© 2016 Author(s). Intraspecific communication is integral to the behavioural ecology of solitary carnivores, but observing and quantifying their communication behaviours in natural environments is difficult. Our systematic literature review found that basic information on scent marking is completely lacking for 23% of all felid species, and information on 21% of other felid species comes solely from one study of captive animals. Here we present results of the first systematic investigation of the scent marking behaviours of Sunda clouded leopards in the wild. Our observations using motion-triggered video cameras in Indonesian Borneo are novel for clouded leopards, and contrary to previous descriptions of their behaviour. We found that clouded leopards displayed 10 distinct communication behaviours, with olfaction, scraping, and cheek rubbing the most frequently recorded. We also showed that males make repeated visits to areas they previously used for marking and that multiple males advertise and receive information at the same sites, potentially enhancing our ability to document and monitor clouded leopard populations. The behaviours we recorded are remarkably similar to those described in other solitary felids, despite tremendous variation in the environments they inhabit, and close a key gap in understanding and interpreting communication behaviours of clouded leopards and other solitary felids.

Sign in / Sign up

Export Citation Format

Share Document