Filters








4,322 Hits in 2.4 sec

Pretraining Reinforcement Learning: Sharpening the Axe Before Cutting the Tree [article]

Saurav Kadavath, Samuel Paradis, Brian Yao
2021 arXiv   pre-print
Pretraining is a common technique in deep learning for increasing performance and reducing training time, with promising experimental results in deep reinforcement learning (RL). However, pretraining requires a relevant dataset for training. In this work, we evaluate the effectiveness of pretraining for RL tasks, with and without distracting backgrounds, using both large, publicly available datasets with minimal relevance, as well as case-by-case generated datasets labeled via self-supervision.
more » ... Results suggest filters learned during training on less relevant datasets render pretraining ineffective, while filters learned during training on the in-distribution datasets reliably reduce RL training time and improve performance after 80k RL training steps. We further investigate, given a limited number of environment steps, how to optimally divide the available steps into pretraining and RL training to maximize RL performance. Our code is available on GitHub
arXiv:2110.02497v1 fatcat:agtvrgfrnrganlngk2elsyvsla

Learning to Localize, Grasp, and Hand Over Unmodified Surgical Needles [article]

Albert Wilcox, Justin Kerr, Brijen Thananjeyan, Jeffrey Ichnowski, Minho Hwang, Samuel Paradis, Danyal Fer, Ken Goldberg
2021 arXiv   pre-print
Robotic Surgical Assistants (RSAs) are commonly used to perform minimally invasive surgeries by expert surgeons. However, long procedures filled with tedious and repetitive tasks such as suturing can lead to surgeon fatigue, motivating the automation of suturing. As visual tracking of a thin reflective needle is extremely challenging, prior work has modified the needle with nonreflective contrasting paint. As a step towards automation of a suturing subtask without modifying the needle, we
more » ... e HOUSTON: Handoff of Unmodified, Surgical, Tool-Obstructed Needles, a problem and algorithm that uses a learned active sensing policy with a stereo camera to localize and align the needle into a visible and accessible pose for the other arm. To compensate for robot positioning and needle perception errors, the algorithm then executes a high-precision grasping motion that uses multiple cameras. In physical experiments using the da Vinci Research Kit (dVRK), HOUSTON successfully passes unmodified surgical needles with a success rate of 96.7% and is able to perform handover sequentially between the arms 32.4 times on average before failure. On needles unseen in training, HOUSTON achieves a success rate of 75 - 92.9%. To our knowledge, this work is the first to study handover of unmodified surgical needles. See https://tinyurl.com/houston-surgery for additional materials.
arXiv:2112.04071v1 fatcat:dtu54auuxvh6tgnletlluuduo4

Lateral Organization in Simulated Four-Component Non-equilibrium Model Membranes

Andrew P. Paradis, Susan R. McKay, Samuel T. Hess
2009 Biophysical Journal  
Paradis 1 , Susan R. McKay 1 , Samuel T. Hess 1,2 . 1 University of Maine, Orono, ME, USA, 2 Institute for Molecular Biophysics, Orono, ME, USA.  ... 
doi:10.1016/j.bpj.2008.12.1396 fatcat:cgqhqguaazdhpmass4zrpbtlti

Applying Depth-Sensing to Automated Surgical Manipulation with a da Vinci Robot [article]

Minho Hwang, Daniel Seita, Brijen Thananjeyan, Jeffrey Ichnowski, Samuel Paradis, Danyal Fer, Thomas Low, Ken Goldberg
2020 arXiv   pre-print
Recent advances in depth-sensing have significantly increased accuracy, resolution, and frame rate, as shown in the 1920x1200 resolution and 13 frames per second Zivid RGBD camera. In this study, we explore the potential of depth sensing for efficient and reliable automation of surgical subtasks. We consider a monochrome (all red) version of the peg transfer task from the Fundamentals of Laparoscopic Surgery training suite implemented with the da Vinci Research Kit (dVRK). We use calibration
more » ... hniques that allow the imprecise, cable-driven da Vinci to reduce error from 4-5 mm to 1-2 mm in the task space. We report experimental results for a handover-free version of the peg transfer task, performing 20 and 5 physical episodes with single- and bilateral-arm setups, respectively. Results over 236 and 49 total block transfer attempts for the single- and bilateral-arm peg transfer cases suggest that reliability can be attained with 86.9 % and 78.0 % for each individual block, with respective block transfer speeds of 10.02 and 5.72 seconds. Supplementary material is available at https://sites.google.com/view/peg-transfer.
arXiv:2002.06302v1 fatcat:wedfovymn5d6dgfli6ll6sro7y

Lipoblastoma in infant: Our experience

Tommaso Armenise, Ottavio Gentile, Antonio Orofino, Samuele Leggio, Maria Paola Lanzillotto, Francesca Zullino, Guglielmo Paradies
2015 Journal of Pediatric Surgery Case Reports  
doi:10.1016/j.epsc.2014.12.001 fatcat:o5yb43773zhnvc5jw2grbhpyla

Pay Attention: Leveraging Sequence Models to Predict the Useful Life of Batteries [article]

Samuel Paradis, Michael Whitmeyer
2019 arXiv   pre-print
We use data on 124 batteries released by Stanford University to first try to solve the binary classification problem of determining if a battery is "good" or "bad" given only the first 5 cycles of data (i.e., will it last longer than a certain threshold of cycles), as well as the prediction problem of determining the exact number of cycles a battery will last given the first 100 cycles of data. We approach the problem from a purely data-driven standpoint, hoping to use deep learning to learn
more » ... patterns in the sequences of data that the Stanford team engineered by hand. For both problems, we used a similar deep network design, that included an optional 1-D convolution, LSTMs, an optional Attention layer, followed by fully connected layers to produce our output. For the classification task, we were able to achieve very competitive results, with validation accuracies above 90%, and a test accuracy of 95%, compared to the 97.5% test accuracy of the current leading model. For the prediction task, we were also able to achieve competitive results, with a test MAPE error of 12.5% as compared with a 9.1% MAPE error achieved by the current leading model (Severson et al. 2019).
arXiv:1910.01347v2 fatcat:ygwt4efdsncynccqfrxvo3wtk4

Intermittent Visual Servoing: Efficiently Learning Policies Robust to Instrument Changes for High-precision Surgical Manipulation [article]

Samuel Paradis, Minho Hwang, Brijen Thananjeyan, Jeffrey Ichnowski, Daniel Seita, Danyal Fer, Thomas Low, Joseph E. Gonzalez, Ken Goldberg
2020 arXiv   pre-print
Correspondence: Samuel Paradis, samparadis@berkeley.edu TABLE I : I Ablation Study Varying Number of Models in Ensemble. TABLE II : II Peg Transfer Baseline Comparison.  ... 
arXiv:2011.06163v1 fatcat:la2pcztinnagngwgfv3bxkw5si

Monitoring and anticipating floods in France

Isabelle Leleu, Bruno Janet, Olivier Piotte, Aurélie Escudier, Céline de Saint-Aubin, Jean-Luc Souldadié, Isabelle Rochet, Didier Jourdan, Denis Paradis, M. Lang, F. Klijn, P. Samuels
2016 E3S Web of Conferences  
doi:10.1051/e3sconf/20160718014 fatcat:qjhdv32wmffrje32kzzlcuk6si

Effects of ribavirin on hepatitis C-associated nephrotic syndrome in four liver transplant recipients

Hung-Patrick Pham, Cyrille Féray, Didier Samuel, Michèle Gigou, Daniel Azoulay, Valérie Paradis, Francis Ducret, Bernard Charpentier, Brigitte Debuire, Antoinette Lemoine
1998 Kidney International  
Effects of ribavirin on hepatitis C-associated nephrotic syndrome in four liver transplant recipients. Background. Hepatitis C virus infection (HCV) is associated with a variety of extrahepatic disorders such as membranoproliferative glomerulonephritis (MPGN), which is generally due to cryoglobulinemia. After liver transplantation for HCV cirrhosis, alpha-interferon treatment against the recurrence of HCV in the liver graft is poorly effective and may induce intractable graft rejection.
more » ... We describe the cases of four liver transplant recipients treated with ribavirin for HCV-related glomerulopathy and nephrotic syndrome. Results. The nephrotic syndrome was attenuated or disappeared during ribavirin therapy, and patients showed a marked decrease in proteinuria and an increase in albuminemia. The syndrome relapsed in two patients when ribavirin therapy was stopped, and a favorable response was again obtained in both cases when the treatment was resumed. The main adverse effect of ribavirin was anemia in two patients with renal impairment. No graft rejection occurred. Conclusions. These findings suggest that continuous therapy with low doses of oral ribavirin may improve the proteinuria of hepatitis C-related glomerulonephritis, at least in liver transplant recipients.
doi:10.1046/j.1523-1755.1998.00113.x pmid:9767549 fatcat:hga6jealtzg6lknynlqf2tu35i

Photometry of the Uranian Satellites with Keck and the Search for Mab

Samuel Paradis, Chris Moeckel, Joshua Tollefson, Imke de Pater
2019 Astronomical Journal  
ORCID IDS Samuel Paradis: 0000-0002-3581-312X Chris Moeckel: 0000-0002-6293-1797 Josh Tollefson: 0000-0003-2344-634X Imke de Pater: 0000-0002-4278-3168  ...  Samuel Paradis was participating through the Undergraduate Research Apprentice Program (URAP) at UC Berkeley.  ... 
doi:10.3847/1538-3881/ab4264 fatcat:zbbfpjkq3baupaoxsey3lwb2n4

Subject realization and crosslinguistic interference in the bilingual acquisition of Spanish and English: what is the role of the input?

JOHANNE PARADIS, SAMUEL NAVARRO
2003 Journal of Child Language  
Paradis & Genesee, 1996) .  ...  For example, neither Paradis & Genesee (1996 nor Hulk & Mü ller (2000) found crosslanguage interference with respect to the use of root infinitives in Germanic/Romance language pairs.  ... 
doi:10.1017/s0305000903005609 fatcat:euorvue5sfhdpd3clqautjqakq

A PROSPECTIVE RANDOMIZED TRIAL OF FK506 VERSUS CYCLOSPORINE AFTER HUMAN PULMONARY TRANSPLANTATION1

BARTLEY P. GRIFFITH, K O BANDO, ROBERT L. HARDESTY, JOHN M. ARMITAGE, ROBERT J. KEENAN, SI M. PHAM, IRVIN L. PARADIS, SAMUEL A. YOUSEM, KANSHI KOMATSU, HIROAKI KONISHI, JOHN J. FUNG, THOMAS E. STARZL
1994 Transplantation  
doi:10.1097/00007890-199403270-00013 pmid:7512292 pmcid:PMC2975521 fatcat:qw64y7hav5gwfpc2tbj6qfinoy

Obliterative bronchiolitis after lung and heart-lung transplantation

Ko Bando, Irvin L. Paradis, Shari Similo, Hiroaki Konishi, Kanshi Komatsu, Thomas G. Zullo, Samuel A. Yousem, John M. Close, Adriana Zeevi, Rene J. Duquesnoy, Jan Manzetti, Robert J. Keenan (+3 others)
1995 Journal of Thoracic and Cardiovascular Surgery  
An analysis of risk factors and management With a prevalence of 34% (55/162 at-risk recipients) and a mortality of 25% (14/55 affected recipients), obliterative bronchiolitis is the most significant long-term complication after pulmonary transplantation. Because of its importance, we examined donor-recipient characteristics and antecedent clinical events to identify factors associated with development of obliterative bronchiolitis, which might be eliminated or modified to decrease its
more » ... . We also compared treatment outcome between recipients whose diagnosis was made early by surveillance transbronchial lung biopsy before symptoms or decline in pulmonary function were present versus recipients whose diagnosis was made later when symptoms or declines in pulmonary function were present. Postoperative airway ischemia, an episode of moderate or severe acute rejection (grade IH/IV), three or more episodes of histologic grade II (or greater) acute rejection, and cytomegalovirus disease were risk factors for development of obliterative bronchiolitis. Recipients with obliterative bronchiolitis detected in the preclinical stage were significantly more likely to be in remission than recipients who had clinical disease at the time of diagnosis: 81% (13/15) versus 33% (13/40); p < 0.05. These results indicate that acute rejection is the most significant risk factor for development of obliterative bronchiolitis and that obliterative bronchiolitis responds to treatment with augmented immunosuppression when it is detected early by surveillance transbronchial biopsy. (J TrIORAC CARDIOVASC SURG 1995;110:4-14)
doi:10.1016/s0022-5223(05)80003-0 pmid:7609567 fatcat:syy7gqbhgncgnjoosqkkiqwqoq

Superhuman Surgical Peg Transfer Using Depth-Sensing and Deep Recurrent Neural Networks [article]

Minho Hwang, Brijen Thananjeyan, Daniel Seita, Jeffrey Ichnowski, Samuel Paradis, Danyal Fer, Thomas Low, Ken Goldberg
2020 arXiv   pre-print
. • Paradis et al. 20 propose intermittent visual servoing (IVS), a paradigm for using a coarse open-loop controller in free space, while using learned visual servoing 21 when the robot is close to  ...  Paradis et al. learn visual servoing from demonstrations, and in peg-transfer experiments, show that the learned model can be robust to tool changes, which would not be the case with the calibration methods  ... 
arXiv:2012.12844v2 fatcat:f5vyqvr3wzh5hbzjdjopbv6s4u

She-Stranger in Paradise: Transnational Gendered Imaginaries in Kurdwin Ayub's Paradise! Paradise! (Paradies! Paradies!, 2016)

Samuel Fernández-Pichel
2019
The present paper examines the construction of overlapping (trans)national and gendered (female) imaginaries in Kurdwin Ayub's Paradise! Paradise! (Paradies! Paradies!, 2016).  ...  She-Stranger in Paradise: Transnational Gendered Imaginaries in Kurdwin … Cinergie. N.16 (2019) its full complexity.  ...  Because, if Paradise! Paradise!  ... 
doi:10.6092/issn.2280-9481/9751 fatcat:zpxhegjptjhuxeaxtgp2hp2gua
« Previous Showing results 1 — 15 out of 4,322 results