Plant recognition by AI: Deep neural nets, transformers, and kNN in deep embeddings

The article reviews and benchmarks machine learning methods for automatic image-based plant species recognition and proposes a novel retrieval-based method for recognition by nearest neighbor classification in a deep embedding space. The image retrieval method relies on a model trained via the Recall@k surrogate loss. State-of-the-art approaches to image classification, based on Convolutional Neural Networks (CNN) and Vision Transformers (ViT), are benchmarked and compared with the proposed image retrieval-based method. The impact of performance-enhancing techniques, e.g., class prior adaptation, image augmentations, learning rate scheduling, and loss functions, is studied. The evaluation is carried out on the PlantCLEF 2017, the ExpertLifeCLEF 2018, and the iNaturalist 2018 Datasets-the largest publicly available datasets for plant recognition. The evaluation of CNN and ViT classifiers shows a gradual improvement in classification accuracy. The current state-of-the-art Vision Transformer model, ViT-Large/16, achieves 91.15% and 83.54% accuracy on the PlantCLEF 2017 and ExpertLifeCLEF 2018 test sets, respectively; the best CNN model (ResNeSt-269e) error rate dropped by 22.91% and 28.34%. Apart from that, additional tricks increased the performance for the ViT-Base/32 by 3.72% on ExpertLifeCLEF 2018 and by 4.67% on PlantCLEF 2017. The retrieval approach achieved superior performance in all measured scenarios with accuracy margins of 0.28%, 4.13%, and 10.25% on ExpertLifeCLEF 2018, PlantCLEF 2017, and iNat2018-Plantae, respectively.

Automatic Fungi Recognition: Deep Learning Meets Mycology

The article presents an AI-based fungi species recognition system for a citizen-science community. The system's real-time identification too - FungiVision - with a mobile application front-end, led to increased public interest in fungi, quadrupling the number of citizens collecting data. FungiVision, deployed with a human-in-the-loop, reaches nearly 93% accuracy. Using the collected data, we developed a novel fine-grained classification dataset - Danish Fungi 2020 (DF20) - with several unique characteristics: species-level labels, a small number of errors, and rich observation metadata. The dataset enables the testing of the ability to improve classification using metadata, e.g., time, location, habitat and substrate, facilitates classifier calibration testing and finally allows the study of the impact of the device settings on the classification performance. The continual flow of labelled data supports improvements of the online recognition system. Finally, we present a novel method for the fungi recognition service, based on a Vision Transformer architecture. Trained on DF20 and exploiting available metadata, it achieves a recognition error that is 46.75% lower than the current system. By providing a stream of labeled data in one direction, and an accuracy increase in the other, the collaboration creates a virtuous cycle helping both communities.

Tracking by Deblatting

Objects moving at high speed along complex trajectories often appear in videos, especially videos of sports. Such objects travel a considerable distance during exposure time of a single frame, and therefore, their position in the frame is not well defined. They appear as semi-transparent streaks due to the motion blur and cannot be reliably tracked by general trackers. We propose a novel approach called Tracking by Deblatting based on the observation that motion blur is directly related to the intra-frame trajectory of an object. Blur is estimated by solving two intertwined inverse problems, blind deblurring and image matting, which we call deblatting. By postprocessing, non-causal Tracking by Deblatting estimates continuous, complete, and accurate object trajectories for the whole sequence. Tracked objects are precisely localized with higher temporal resolution than by conventional trackers. Energy minimization by dynamic programming is used to detect abrupt changes of motion, called bounces. High-order polynomials are then fitted to smooth trajectory segments between bounces. The output is a continuous trajectory function that assigns location for every real-valued time stamp from zero to the number of frames. The proposed algorithm was evaluated on a newly created dataset of videos from a high-speed camera using a novel Trajectory-IoU metric that generalizes the traditional Intersection over Union and measures the accuracy of the intra-frame trajectory. The proposed method outperforms the baselines both in recall and trajectory accuracy. Additionally, we show that from the trajectory function precise physical calculations are possible, such as radius, gravity, and sub-frame object velocity. Velocity estimation is compared to the high-speed camera measurements and radars. Results show high performance of the proposed method in terms of Trajectory-IoU, recall, and velocity estimation.

Fine-grained recognition of plants from images

BACKGROUND

Fine-grained recognition of plants from images is a challenging computer vision task, due to the diverse appearance and complex structure of plants, high intra-class variability and small inter-class differences. We review the state-of-the-art and discuss plant recognition tasks, from identification of plants from specific plant organs to general plant recognition "in the wild".

RESULTS

We propose texture analysis and deep learning methods for different plant recognition tasks. The methods are evaluated and compared them to the state-of-the-art. Texture analysis is only applied to images with unambiguous segmentation (bark and leaf recognition), whereas CNNs are only applied when sufficiently large datasets are available. The results provide an insight in the complexity of different plant recognition tasks. The proposed methods outperform the state-of-the-art in leaf and bark classification and achieve very competitive results in plant recognition "in the wild".

CONCLUSIONS

The results suggest that recognition of segmented leaves is practically a solved problem, when high volumes of training data are available. The generality and higher capacity of state-of-the-art CNNs makes them suitable for plant recognition "in the wild" where the views on plant organs or plants vary significantly and the difficulty is increased by occlusions and background clutter.

Rotation-invariant image and video description with local binary pattern features

In this paper, we propose a novel approach to compute rotation-invariant features from histograms of local noninvariant patterns. We apply this approach to both static and dynamic local binary pattern (LBP) descriptors. For static-texture description, we present LBP histogram Fourier (LBP-HF) features, and for dynamic-texture recognition, we present two rotation-invariant descriptors computed from the LBPs from three orthogonal planes (LBP-TOP) features in the spatiotemporal domain. LBP-HF is a novel rotation-invariant image descriptor computed from discrete Fourier transforms of LBP histograms. The approach can be also generalized to embed any uniform features into this framework, and combining the supplementary information, e.g., sign and magnitude components of the LBP, together can improve the description ability. Moreover, two variants of rotation-invariant descriptors are proposed to the LBP-TOP, which is an effective descriptor for dynamic-texture recognition, as shown by its recent success in different application problems, but it is not rotation invariant. In the experiments, it is shown that the LBP-HF and its extensions outperform noninvariant and earlier versions of the rotation-invariant LBP in the rotation-invariant texture classification. In experiments on two dynamic-texture databases with rotations or view variations, the proposed video features can effectively deal with rotation variations of dynamic textures (DTs). They also are robust with respect to changes in viewpoint, outperforming recent methods proposed for view-invariant recognition of DTs.

Rotation-invariant image and video description with local binary pattern features

In this paper, we propose a novel approach to compute rotation-invariant features from histograms of local noninvariant patterns. We apply this approach to both static and dynamic local binary pattern (LBP) descriptors. For static-texture description, we present LBP histogram Fourier (LBP-HF) features, and for dynamic-texture recognition, we present two rotation-invariant descriptors computed from the LBPs from three orthogonal planes (LBP-TOP) features in the spatiotemporal domain. LBP-HF is a novel rotation-invariant image descriptor computed from discrete Fourier transforms of LBP histograms. The approach can be also generalized to embed any uniform features into this framework, and combining the supplementary information, e.g., sign and magnitude components of the LBP, together can improve the description ability. Moreover, two variants of rotation-invariant descriptors are proposed to the LBP-TOP, which is an effective descriptor for dynamic-texture recognition, as shown by its recent success in different application problems, but it is not rotation invariant. In the experiments, it is shown that the LBP-HF and its extensions outperform noninvariant and earlier versions of the rotation-invariant LBP in the rotation-invariant texture classification. In experiments on two dynamic-texture databases with rotations or view variations, the proposed video features can effectively deal with rotation variations of dynamic textures (DTs). They also are robust with respect to changes in viewpoint, outperforming recent methods proposed for view-invariant recognition of DTs.

Feature-based affine-Invariant localization of faces

We present a novel method for localizing faces in person identification scenarios. Such scenarios involve high resolution images of frontal faces. The proposed algorithm does not require color, copes well in cluttered backgrounds, and accurately localizes faces including eye centers. An extensive analysis and a performance evaluation on the XM2VTS database and on the realistic BioID and BANCA face databases is presented. We show that the algorithm has precision superior to reference methods.

Serum fructosamine is not a useful screening test for gestational diabetes

Serum fructosamine was measured in 569 samples of pregnant women without gestational diabetes. We defined abnormal fructosamine as mean + 2SD, and analysed its potential value to detect patients with gestational diabetes diagnosed with current screening criteria. We found serum fructosamine to be an insensitive parameter: Measured at the time of a positive 50 g glucose screening, SF would have detected 4/48 gestational diabetes.

Solubilization of microsomal-associated phosphatidylserine synthase and phosphatidylinositol synthase from Saccharomyces cerevisiae

Membrane-associated cytidine 5'-diphospho-1,2-diacyl-sn-glycerol (CDP-diacylglycerol):L-serine O-phosphatidyltransferase (phosphatidylserine synthase, EC2.7.8.8.) and CDP-diacylglycerol:myo-inositol phosphatidyltransferase (phosphatidylinositol synthase, EC 2.7.8.11) were solubilized from the microsomal fraction of Saccharomyces cerevisiae. A variety of detergents were examined for their ability to release phosphatidylserine synthase and phosphatidylinositol synthase activities from the microsome fraction. Both enzymes were solubilized from the microsome fraction with Renex 690 in yield over 80% with increase to specific activity of 1.6-fold. Both solubilized enzymatic activities were dependent on manganese ions and Triton X-100 for maximum activity. The pH optimum for each reaction was 8.0. The apparent Km values for CDP-diacylglycerol and serine for the phosphatidylserine synthase reaction were 0.1 and 0.25 mM, respectively. The apparent Km values for CDP-diacylglycerol and inositol for the phosphatidylinositol synthase reaction were 70 microM and 0.1 mM, respectively. Thioreactive agents inhibited both enzymatic activities. Both solubilized enzymatic activities were thermally inactivated at temperatures above 30 degrees C.