We propose a method for lossy image compression based on recurrent, convolutional neural networks that outperforms BPG (4:2:0), WebP, JPEG2000, and JPEG as measured by MS-SSIM. We introduce three improvements over previous research that lead to this state-of-the-art result. First, we show that training with a pixel-wise loss weighted by SSIM increases reconstruction quality according to several metrics. Second, we modify the recurrent architecture to improve spatial diffusion, which allows the network to more effectively capture and propagate image information through the network's hidden state. Finally, in addition to lossless entropy coding, we use a spatially adaptive bit allocation algorithm to more efficiently use the limited number of bits to encode visually complex image regions. We evaluate our method on the Kodak and Tecnick image sets and compare against standard codecs as well recently published methods based on deep neural networks.

In learning-based approaches to image compression, codecs are developed by optimizing a computational model to minimize a rate-distortion objective. Currently, the most effective learned image codecs take the form of an entropy-constrained autoencoder with an entropy model that uses both forward and backward adaptation. Forward adaptation makes use of side information and can be efficiently integrated into a deep neural network. In contrast, backward adaptation typically makes predictions based on the causal context of each symbol, which requires serial processing that prevents efficient GPU / TPU utilization. We introduce two enhancements, channel-conditioning and latent residual prediction, that lead to network architectures with better rate-distortion performance than existing context-adaptive models while minimizing serial processing. Empirically, we see an average rate savings of 6.7% on the Kodak image set and 11.4% on the Tecnick image set compared to a context-adaptive baseline model. At low bit rates, where the improvements are most effective, our model saves up to 18% over the baseline and outperforms hand-engineered codecs like BPG by up to 25%.

P Pu ur rp po os se e: : To evaluate the comparative preemptive effects of gabapentin and tramadol on postoperative pain and fentanyl requirement in laparoscopic cholecystectomy.M Me et th ho od ds s: : Four hundred fifty-nine ASA I and II patients were randomly assigned to receive 300 mg gabapentin, 100 mg tramadol or placebo in a double-blind manner two hours before laparoscopic cholecystectomy under general anesthesia. Postoperatively, patients' pain scores were recorded on a visual analogue scale every two hours for the initial 12 hr and thereafter every three hours for the next 12 hr. Patients received fentanyl 2 µg·kg -1 intravenously on demand. The total fentanyl consumption for each patient was recorded.R Re es su ul lt ts s: : Patients in the gabapentin group had significantly lower pain scores at all time intervals (2.65 ± 3.00, 1.99 ± 1.48, 1.40 ± 0.95, 0.65 ± 0.61) in comparison to tramadol (2.97 ± 2.35, 2.37 ± 1.45, 1.89 ± 1.16, 0.87 ± 0.50) and placebo (5.53 ± 2.22, 3.33 ± 1.37, 2.41 ± 1.19, 1.19 ± 0.56). Significantly less fentanyl was consumed in the gabapentin group (221.16 ± 52.39 µg) than in the tramadol (269.60 ± 44.17 µg) and placebo groups (355.86 ± 42.04 µg; P < 0.05). Sedation (33.98%), nausea/retching/vomiting (24.8%) were the commonest side effects in the gabapentin group whereas respiratory depression (3.9%) was the commonest in the tramadol group and vertigo (7.8%) in the placebo group. C Co on nc cl lu us si io on n: : Preemptive use of gabapentin significantly decreases postoperative pain and rescue analgesic requirement in laparoscopic cholecystectomy. (2,65 ± 3,00; 1,99 ± 1,48; 1,40 ± 0,95; 0,65 ± 0,61) que ceux du groupe tramadol (2,97 ± 2,35; 2,37 ± 1,45; 1,89 ± 1,16; 0,87 ± 0,50) ou placebo (5,53 ± 2,22; 3,33 ± 1,37; 2,41 ± 1,19; 1,19 ± 0,56). La demande de fentanyl a été significativement plus basse avec la gabapentine (221,16 ± 52,39 µg) qu'avec le tramadol (269,60 ± 44,17 µg) ou le placebo (355,86 ± 42,04 µg; P < 0,05). La sédation (33,98 %), les nausées/haut-lecoeur/vomissements (24,8 %) ERIPHERAL tissue injury provokes peripheral sensitization (a reduction in the threshold of nociceptor afferent peripheral terminals) and central sensitization (an activity dependent increase in the excitability of spinal neurons). 1,2 These changes contribute to the postinjury pain hypersensitivity state which manifests as an increase in the responsiveness to noxious stimuli and a decrease in the pain threshold, both at the site of injury and in the surrounding uninjured tissue. 1,2 The optimal form of treatment is that applied pre, intra and postoperatively to preempt the establishment of pain hypersensitivity during and after surgery. The preemptive treatment could be directed at the periphery, at inputs along sensory axons, and at central neurons. Different treatment regimens could be used at different times relative to surgery to maximize the prevention of pain in response to different levels of sensory inputs. 1,2 Gabapentin and tramadol both have demonstrated analgesic effect...

The leading approach for image compression with artificial neural networks (ANNs) is to learn a nonlinear transform and a fixed entropy model that are optimized for rate-distortion performance. We show that this approach can be significantly improved by incorporating spatially local, image-dependent entropy models. The key insight is that existing ANN-based methods learn an entropy model that is shared between the encoder and decoder, but they do not transmit any side information that would allow the model to adapt to the structure of a specific image. We present a method for augmenting ANN-based image coders with image-dependent side information that leads to a 17.8% rate reduction over a state-of-theart ANN-based baseline model on a standard evaluation set, and 70-98% reductions on images with low visual complexity that are poorly captured by a fixed, global entropy model.

We review a class of methods that can be collected under the name nonlinear transform coding (NTC), which over the past few years have become competitive with the best linear transform codecs for images, and have superseded them in terms of rate-distortion performance under established perceptual quality metrics such as MS-SSIM. We assess the empirical rate-distortion performance of NTC with the help of simple example sources, for which the optimal performance of a vector quantizer is easier to estimate than with natural data sources. To this end, we introduce a novel variant of entropy-constrained vector quantization. We provide an analysis of various forms of stochastic optimization techniques for NTC models; review architectures of transforms based on artificial neural networks, as well as learned entropy models; and provide a direct comparison of a number of methods to parameterize the rate-distortion trade-off of nonlinear transforms, introducing a simplified one.