Deep neural network (DNN) quantization converting floating-point (FP) data in the network to integers (INT) is an effective way to shrink the model size for memory saving and simplify the operations for compute acceleration. Recently, researches on DNN quantization develop from inference to training, laying a foundation for the online training on accelerators. However, existing schemes leaving batch normalization (BN) untouched during training are mostly incomplete quantization that still adopts high precision FP in some parts of the data paths. Currently, there is no solution that can use only low bit-width INT data during the whole training process of largescale DNNs with acceptable accuracy. In this work, through decomposing all the computation steps in DNNs and fusing three special quantization functions to satisfy the different precision requirements, we propose a unified complete quantization framework termed as "WAGEUBN" to quantize DNNs involving all data paths including W (Weights), A (Activation), G (Gradient), E (Error), U (Update), and BN. Moreover, the Momentum optimizer is also quantized to realize a completely quantized framework. Experiments on ResNet18/34/50 models demonstrate that WAGEUBN can achieve competitive accuracy on ImageNet dataset. For the first time, the study of quantization in largescale DNNs is advanced to the full 8-bit INT level. In this way, all the operations in the training and inference can be bit-wise operations, pushing towards faster processing speed, decreased memory cost, and higher energy efficiency. Our throughout quantization framework has great potential for future efficient portable devices with online learning ability.

No abstract

Few-shot learning has recently emerged as a new challenge in the deep learning field: unlike conventional methods that train the deep neural networks (DNNs) with a large number of labeled data, it asks for the generalization of DNNs on new classes with few annotated samples. Recent advances in few-shot learning mainly focus on image classification while in this paper we focus on object detection. The initial explorations in few-shot object detection tend to simulate a classification scenario by using the positive proposals in images with respect to certain object class while discarding the negative proposals of that class. Negatives, especially hard negatives, however, are essential to the embedding space learning in few-shot object detection. In this paper, we restore the negative information in few-shot object detection by introducing a new negative-and positive-representative based metric learning framework and a new inference scheme with negative and positive representatives. We build our work on a recent few-shot pipeline RepMet [1] with several new modules to encode negative information for both training and testing. Extensive experiments on ImageNet-LOC and PASCAL VOC show our method substantially improves the state-of-the-art few-shot object detection solutions. Our code is available at https://github.com/yang-yk/NP-RepMet.

No abstract

Background: To compare the intraoperative and postoperative outcomes of indocyanine green (ICG) administration in robot-assisted partial nephrectomy (RAPN) and report the differences in the results between patients with benign and malignant renal tumors. Methods: From 2017 to 2020, 132 patients underwent RAPN at our institution, including 21 patients with ICG administration. Clinical data obtained from our institution’s RAPN database were retrospectively reviewed. Intraoperative, postoperative, pathological, and functional outcomes of RAPN were assessed. Results: The pathological results indicated that among the 127 patients, 38 and 89 had received diagnoses of benign and malignant tumors, respectively. A longer operative time (311 vs. 271 min; p = 0.006) but superior preservation of estimated glomerular filtration rate (eGFR) at 3-month follow-up (90% vs. 85%; p = 0.031) were observed in the ICG-RAPN group. Less estimated blood loss, shorter warm ischemia time, and superior preservation of eGFR at postoperative day 1 and 6-month follow-up were also noted, despite no significant differences. Among the patients with malignant tumors, less estimated blood loss (30 vs. 100 mL; p < 0.001) was reported in the ICG-RAPN subgroup. Conclusions: Patients with ICG-RAPN exhibited superior short-term renal function outcomes compared with the standard RAPN group. Of the patients with malignant tumors, ICG-RAPN was associated with less blood loss than standard RAPN without a more positive margin rate. Further studies with larger cohorts and prospective designs are necessary to verify the intraoperative and functional advantages of the green dye.