भाल जब

Multi-scale modeling: Math 646

Multi-phase phenomena: ME/ChE 632

Three Mandatory Courses

Attendance is required at seminars offered by the program

Recommendation by student’s major professor and by a faculty member in the program.  Graduate research that requires a significant degree of scientific computing and is an appropriate area of science and engineering.

At least one faculty participant in ASC (e.g. teaching one of the courses below) will be on the Program of Study committee of any student in the program.  That person will be in a department different from the student’s major.

The heads are then concatenated and transformed using a square weight matrix WO∈Rdmodel×dmodel\textbf{W}^{O} \in R^{d_{model} \times d_{model}}WO∈Rdmodel​×dmodel​, since dmodel=hdkd_{model}=h d_{k}dmodel​=hdk​

Lil'Log

TEM instruments have multiple operating modes including conventional imaging, scanning TEM imaging (STEM), diffraction, spectroscopy, and combinations of these. Even within conventional imaging, there are many fundamentally different ways that contrast is produced, called "image contrast mechanisms". Contrast can arise from position-to-position differences in the thickness or density ("mass-thickness contrast"), atomic number ("Z contrast", referring to the common abbreviation Z for atomic number), crystal structure or orientation ("crystallographic contrast" or "diffraction contrast"), the slight quantum-mechanical phase shifts that individual atoms produce in electrons that pass through them ("phase contrast"), the energy lost by electrons on passing through the sample ("spectrum imaging") and more. Each mechanism tells the user a different kind of information, depending not only on the contrast mechanism but on how the microscope is used—the settings of lenses, apertures, and detectors. What this means is that a TEM is capable of returning an extraordinary variety of nanometer- and atomic-resolution information, in ideal cases revealing not only where all the atoms are but what kinds of atoms they are and how they are bonded to each other. For this reason TEM is regarded as an essential tool for nanoscience in both biological and materials fields.

Transmission electron microscopes are capable of imaging at a significantly higher resolution than light microscopes, owing to the smaller de Broglie wavelength of electrons. This enables the instrument to capture fine detail—even as small as a single column of atoms, which is thousands of times smaller than a resolvable object seen in a light microscope. Transmission electron microscopy is a major analytical method in the physical, chemical and biological sciences. TEMs find application in cancer research, virology, and materials science as well as pollution, nanotechnology and semiconductor research, but also in other fields such as paleontology and palynology.

Transmission electron microscopy (TEM) is a microscopy technique in which a beam of electrons is transmitted through a specimen to form an image. The specimen is most often an ultrathin section less than 100 nm thick or a suspension on a grid. An image is formed from the interaction of the electrons with the sample as the beam is transmitted through the specimen. The image is then magnified and focused onto an imaging device, such as a fluorescent screen, a layer of photographic film, or a sensor such as a scintillator attached to a charge-coupled device.

In contrast to traditional siliconbased solar cells, organic (or polymer) solar cells are low weight,printable on flexible substrate, and most importantly, can be pro-duced at room temperature at very low cost.

The principal process parameters that influence OPV performance include the annealing time and temperature, blend ratio of the donor to the acceptor material, and nature of the solvent.4,5

In one of the popular methods of device fabrication, the donor (D) and the acceptor (A) are dissolved in a suitable solvent and then spin-coated onto the substrate. During spin-coating, the solvent evaporates, resulting in the D–A materials spontaneously phase-transforming to D and A domains within the active organic-semiconductor layer.

However, these advantages are offset by the challenges faced due to low stability, low efficiency, and process sensitivity.

Fabrication of OPVs is in contrast to the energy-intensive production of their inorganic counterparts. Therefore, OPVs have the potential for high energy return on investment. Another advantage of organic materials is that they have a high optical absorption coefficient. Therefore, thin films can capture sufficient solar energy, thus reducing material cost.

In other words, quantitative characterization of the microstructure is critical for understanding device performance

Organic electronics (electronic devices made from polymers or similar organic compounds) are an important class of thin film devices that exhibit flexibility, stretchability, softness and compatibility with biological systems, thus making them exceedingly attractive for bioelectronics and healthcare applications. Performance of these devices is highly correlated with the microstructure within the thin film.

While several microscopy based imaging techniques – including Transmission Electron Microscopy (TEM) – are currently available to visualize the microstructure, analyzing the resulting data has remained predominantly manual. This makes extracting microstructural information from microscopic images subjective, and slow.

We introduce a mathematical framework and associated software, GRATE, that integrates concepts from graph theory, image analysis and hierarchical (quad-tree) image decomposition to automatically extract a suite of microstructural features from TEM image data. Both local traits like lattice spacing, orientation and size measures as well as global traits like number and size distribution of crystallites are efficiently extracted. The software is modular, extensible and is packaged into a simple graphical user interface that can reduce time of adoption by practitioners.

The fighters of the Mughal centre including Madhav Singh soon raced to the rescue of their commander, Badayuni mentioning the Rana squaring off against Madhav Singh and ‘a shower of arrows being poured on him’. It is here that the redoubtable Raja Ramshah Tomar ‘repaid the debt of gratitude to Mewar with his life-blood’. Continually shielding the Rana from frontal attacks, and showing ‘extreme obstinacy of resistance’,

Those of the army who had fled on the first attack, did not draw rein till they had passed five or six crosses beyond the river.

The other division of Ráná Kíká’s army, under the Ráná in person, charged out of the pass, and meeting Qází Khán, who was at the entrance of the pass, swept his men before them, and bearing them along broke through his centre. Then the Shaikh-sons from Síkrí all fled at once. And an arrow struck Shaikh Mānsur, (son-in-law of Shaikh Ibráhím) who was the leader of this company, in the seat of honour as he was in the act of flight, and he bore the wound for a considerable time. But Qází Khán, although he was but a Mulla, stood his ground manfully, until receiving a scimitar blow on his right hand, which wounded his thumb, being no longer able to hold his own, he recited [the saying] ‘Flight from overwhelming odds is one of the traditions of the prophet,’and followed his men [in their retreat].

A treacherous scene then played out, Mān Singh perhaps watching helplessly from behind. With the skirmishers under Hashim Barha and the Mughal advance body having hundreds of Rajputs being shoved back in close combat with the Mewar army, Asaf Khan and Badayuni from close behind could not distinguish anymore between friendly and hostile Rajputs, and aiming for the right one became a question. Writes Badayuni. At this juncture, the author, who was with some of the special troops of the advance-body said to Ásaf, “How are we now in these circumstances to distinguish between friendly and hostile Rájpúts?” He answered, “They will experience the whiz of the arrows, be what may, on whichever side they may belong, their being killed will be a gain to Islám. So we kept firing away, and our aim at such a mountain-like mass of men never missed. And it became certain that my hand prospered in the matter, and that I attained the reward due to one who fights against infidels.” Thus many of the Rajputs fighting from the Mughal side were slain by their own army commanders.

The skirmishers, unable to bear the onslaught, fled helter-skelter back onto the advance body, leading to a commotion that went out of control. Badayuni describes in vivid detail “And the Rájpúts of our army, the leader of whom was Rájah Loun Karan, and who were most of them on the left, ran away like a flock of sheep, and breaking through the ranks of the advance-body fled for protection to our right wing”.

about an hour after daybreak, they met! The ‘Chickens of the front-line’ under Hashim Barha, who had approached westward of the pass, were met with a lightning-like charge of the smaller Mewar division coming from the west of the pass and were hurled back towards the Mughal advance body a few hundred meters behind them, sustaining a complete rout.

Rana Pratap strategised to wait until the Mughal army crossed the river and moved right up to where the undulating terrain starts and extends till the mouth of the Haldighati. He then divided his cavalry of three thousand into two divisions, the smaller one led by Hakim Khan Sur and warriors such as Ramdass Rathore, Krishnadas Chundawat and Rawat Netsi. The main division, commanded by Pratap himself had Raja Ramshah Tomar in the right wing and Jhala Mān Singh commanding the left wing. Barhats Jaisa and Keshav are believed to have led the rear of the Mewar army.

The Mughal formation was classical, imaginable as an arrow-head- Mān Singh was seated on his elephant in the centre surrounded by cavalry led by Madho Singh and other Rajputs, Muhammad Rafi Badakshi,  Murad Uzbek and other ‘youths of renown’ as Badayuni describes them. The left-wing was led by Qazi Khan and the Shaikhzadas of Sikri, the right-wing was led, as per tradition, by the Saiyyads of Barha commanded by Ahmad Khan Barha. The rear of the Mughal formation, which eventually played a critical role, was led by Mihtar Khan. The ‘vanguard’ or advance body in front of the centre was manned by Asaf Khan, Jagannath Kachhwa and Ghiasuddin Ali and in front of this advance body was sent a body of 80 soldiers under Hashim Barha which Badayuni fittingly calls chuza-i-harawwal, translated by Lowe as ‘the chickens of the frontline’. Badayuni was present with Asaf Khan in the van.

On that day of days, both camps were astir well before sunrise, and cautiously the Mughal forces started their advance from Molela south and south-west crossing the Banas river while the Mewar army a few miles away manoeuvred to inflict maximum damage on their foes, who outnumbered them 2 to 1. The theatre of the battle is most likely to have been between the upper end of the Haldighati gorge which runs east and north-east, where the battle started and the open grounds around the village of Khamnor to the north of the pass where it moved to, which itself is bordered on the northern side by the Banas River and to the south/south-west of which lies the Haldighati pass

To counter this imminent onslaught the Mewar forces prepared at their base in Gogunda(called Kokandah by Badayuni), about 20 miles from Haldighati- a narrow gorge so-called since the rock-sand at the place is of such colour, as to appear to an onlooker to be covered with turmeric

Ghazi Khan Badakshi, Syed Ahmad Barha, Syed Hashim Barha, Qazi Khan, the Shaikhzadas of Sikri, Mihtar Khan, Ali Murad Uzbek and Rajputs such as Lonkaran Kachhwa, Madhav Singh and Jagannath Kachhwa were commanders of various units of the invading army

Rana Pratap is called Rana Kika in Mughal records, as in many records of the period. Ranas of Mewar were often called by their nicknames. Sangram Singh was Sanga, Kumbhkaran Singh was Kumbha. Pratap was thus Kika

The failure of parleys and the defiance of the Rana made Akbar decide on war. Holding the Mughal court at Ajmer, he chose Mān Singh whom he considered one of the ablest and bravest among his grandees to lead an army to force Mewar into submission. He was entrusted with this responsibility at the shrine of Khwaja Moinuddin Chisti, where Akbar, “treating him with kindness and showing him the greatest favour, presented him with a robe of honour and a horse with all its appointments, and ordered him to proceed to the hostile district of Gogoonda and Kumbhalmer which was a dependency belonging to Rana Kika” as Badayuni records.

Peace was pursued by both sides in repeated parleys in years preceding the battle- through Mughal envoys such as Raja Bhagwan Das, Todal Mal and Jalal Khan meeting the Mewaris, but inevitably ending in a deadlock for Maharana Pratap although unfailingly courteous and hospitable, was resolutely disinclined to accepting Mughal supremacy. Perhaps the most famous of episodes in such parleys as preserved in folklore is that of Mān Singh meeting Maharana Pratap with a proposal for an alliance on the banks of the lake Udaisagar. The rulers of Amber had until recently theretofore been vassals of Mewar, and at the feast that was laid out for the parley Pratap would not condescend to treating Mān as an equal, and instead sent his son Amar Singh to join in at the ceremony, expressing his regret under the pretext of being indisposed. Mān demanded in no uncertain terms that Pratap himself join the feast, to which the Rana responded in the negative.  This and other rancorous exchange led to Mān leaving the feast abruptly, with the ominous warning ‘’Abide then in peril, if such be your resolve; but remember- I shall come again and if I do not humble your pride, my name is not Mān!’’ as Tod records, to which the Rana replied ‘’he should always be happy to meet him’’ while someone-Dodiya Bhim according to Kesri Singh- added in plainer parlance desiring Mān not to forget to bring his ‘Phoopha’ (father’s sister’s husband), Akbar, along!

The fortunes of the Mughals and the Mewaris had swung inversely proportional in the decades preceding the battle of Haldighati. While the Mughals had risen from being rulers who had to flee their ancestral lands in strife to getting firmly entrenched in and thereafter constantly expanding their suzerainty over large parts of northern, western and central India, Mewar had reduced from being the leader of Rajput confederacy in India to one left fighting to retain control of its independence while getting encircled by subjugated provinces brought under the Mughal fold.

Few battles of Indian history have captivated the collective psyche of modern India as the Battle of Haldighati, fought from dawn till noon on a hot summer day of the 18th of June, 1576 between the hegemony-seeking Mughal forces of Akbar, then 34, led by the 26-year-old Kuwar Mān Singh of Amber and the forces of Mewar striving to preserve their freedom and independence, led by Maharana Pratap Singh, aged 36.

Mewar folklore describes, concurrent to this engagement, Pratap’s famous duel with the burly warrior Bahlol Khan, ending with the heavy sword of Pratap crashing ‘like lightening on to Bahlol’s headpiece cleaving the Mughal vertically and disembowelling his steed as well’.

War elephants from both sides then joined in the fight. Jamal Khan Faujdar brought forward the elephant Gajmukta, which was countered by Mewar’s ‘rank-breaking’ Lona. The Mewar hulk dominated and grievously wounded the Mughal elephant, which was about to fly when a bullet struck and killed the Mahawat of Lona, and rider-less it wandered out of battle. Just then, the most famous elephant of the Mewar force, Ram Prasad, which had been the subject of discussion and wonder several times in Akbar’s court, and which Rana Pratap repeatedly declined to submit to the collection of the Mughal King was brought forward by Pratap Singh, youngest son of Raja Ramshah Tomar and it took down several men of the opposing force. Amidst the disarray caused in the Mughal ranks by its vicious attacks, Kamal Khan brought forward the elephant Gajraj and Panjoo brought forward Ran Mandar- both these elephants got simultaneously dominated by Ram Prasad and in the words of Abu Fazl, Ran Mandar too ‘was letting the foot of his courage slip’, when perchance an arrow killed the Mahawat of the elephant, and it too became driver-less. Just then Panjoo, in Badayuni’s words ‘leapt from his own elephant, and took his seat on that of the Rana, and performed such a deed as none other could have’. The famed Ram Prasad thus came under Mughal control, becoming the only war-trophy for the Mughals from the battle.

Once the battle moved back a few hundred metres from the uplands to the plains around Khamnor, ever-since called Rakht-tal (field of Blood) the clash continued in right earnest. In the words of Abu Fazl, ‘the price of life was low, that of honour high’. Although the savage onslaught of the Mewar army did throw the Mughal forces into complete disarray, the Sayyids of Barha on the right resolutely held their ground despite the various fleeing formations racing to them for protection and avoided an out-and-out Mughal defeat. Says Badayuni “if the Sayyids had not held their ground firmly, such confusion did the retreating advance-body cause in their ranks, that the affair would have turned out a disgraceful defeat.” Shortly thereafter, Badayuni himself fled, blaming the retreating Rajputs of Mān Singh ‘for the flight of Asaf Khan’ and that of ‘our worthy author’ as translator Lowe gently put it. Seeing the rout of the advance body and the left wing, Mān Singh hastened to join the fight, and it was in the nick of time that his Altamsh arrived and saved the life of Jagannath Kachhwa. Records Abu fazl: “Jagannāth behaved bravely and was about to sacrifice his life when the Altamsh arrived, and Kuar Mān Singh in person joined in the fight.” Ramdas Rathore in the Mewar van fought in hand to hand combat and lost his life to Jagannath Kachhwa.

An important leader of the Mewar army was Hakim Khan Sur- a descendant of the Sher Shah Suri family who nursed a hereditary grudge against the Mughals. Hakim Khan was most probably a free-lancer who joined the Mewar army due to a common cause, for there is no mention of him prior to or after this battle in any record

The various Rajput chieftains who had rallied to his cause included Jhala Mān Singh of Badi Sadri, Krishnadas Chundawat of Salumbar, Rawat Netsi and his son Mān Singh, Jhala Mān Singh Sajjawat of Delwara, Dodiya Bhim Singh of Lava, Raja Ramshah Tomar of Gwalior (grandson of the famous Raja Mān Singh Tomar) and his sons Salivahan Singh, Bhan Singh and Pratap Singh, Rathor Sankar Das and his son Narhari Das and two brothers Keindas and Ramdas, Rathore Ramdas of Badnor (son of the famous defender of the Chittor fort, Jaimal), the Charans Rama Sandu, Barhat Jaisa and Barhat Keshav of Soniyana and the Bhil tribal chief Rao Poonja.

(log(D(x))+log(D(G(z)))log(D(x))+log(D(G(z)))log(D(x)) + log(D(G(z)))).

referred to by that checksum

checksummed

cmake .. -DCMAKE_BUILD_TYPE=Release -DENABLE_4D=Yes

cmake .. -DCMAKE_BUILD_TYPE=Release -DENABLE_3D=Yes

cmake .. -DCMAKE_BUILD_TYPE=Release -DENABLE_2D=Yes

The explosion of image data on the Internet has the potential to foster more sophisticated and robust models and algorithms to index, retrieve, organize and interact with images and multimedia data. But exactly how such data can be harnessed and organized remains a critical problem. We introduce here a new database called “ImageNet”, a large-scale ontology of images built upon the backbone of the WordNet structure. ImageNet aims to populate the majority of the 80,000 synsets of WordNet with an average of 500-1000 clean and full resolution images. This will result in tens of millions of annotated images organized by the semantic hierarchy of WordNet. This paper offers a detailed analysis of ImageNet in its current state: 12 subtrees with 5247 synsets and 3.2 million images in total. We show that ImageNet is much larger in scale and diversity and much more accurate than the current image datasets. Constructing such a large-scale database is a challenging task. We describe the data collection scheme with Amazon Mechanical Turk. Lastly, we illustrate the usefulness of ImageNet through three simple applications in object recognition, image classification and automatic object clustering. We hope that the scale, accuracy, diversity and hierarchical structure of ImageNet can offer unparalleled opportunities to researchers in the computer vision community and beyond.

A simulator for X-ray images is presented based on a virtual X-ray source and a virtual human body obtained from tomographic slices. In the simulator it is possible to modify the tube potential, the anodic current, the exposure time, the filtration and some geometric parameters such as source–skin distance, orientation and field size. The virtual body consists of a three-dimensional voxel matrix in which CT numbers for each point of the body are stored. The interactions of X rays passing through the body are evaluated using the pencil beam technique. The image is obtained by computing the dose absorbed by the detector and converting it into optical density using a proper response function. The image spatial resolution is limited by the voxel size. The influence of each parameter on the image quality can be observed interactively. The dose absorbed in each point of the body is an important parameter obtained as output of the simulator.

Other approaches to this kind of simulation exist. They differ from ours either in the use of different methods such as the Monte Carlo technique, following the path of each photon(1,2), or the deterministic approach, based on the integral photon transport equation(3) or using as objects to be imaged computer-aided drawing (CAD) models

The system presented in this paper consists of three virtual objects, the X-ray source, the patient's body and the X-ray detector. The X-ray source is modelled in such a way as to reproduce the energy spectrum emitted by an X-ray tube. The virtual body is a 3-D representation obtained from the processing of a series of CT slices. The detector is modelled according to its physical characteristics and its response function. The beam is segmented both in energy and in solid angle (pencil beam) and its attenuation is computed along its path.

The work presented here is part of a more general project named MOCAMA (MOnte CArlo MAchine), which foresees the realisation of dedicated hardware based on programmable logic circuits (FPGA: Field Programmable Gate Array) for solving physics problems requiring long computation time, such as the compilation of treatment planning in radiotherapy, showing the dose distribution map for cancer treatment. In the framework of this project, a set of software programs is being developed for different tasks. Some examples are the three-dimensional (3-D) visualisation of surfaces of the same density obtained from CT slices, the simulation of phantoms irradiated in different conditions and the simulation of a radiological system, which is the subject of the present paper.

tar -xzvf cudnn-7.0-linux-x64-v4.0-prod.tgz sudo cp cuda/lib64/libcudnn* /usr/local/cuda-7.0/lib64/ sudo cp cuda/include/cudnn.h /usr/local/cuda-7.0/include/

luarocks install cudnn

th -e "require 'cutorch'; require 'cunn'; print(cutorch)"

luarocks install cutorch luarocks install cunn

The master branch of cuDNN.torch does not support cuDNN v7. Installing from R7 branch probably works fine. $ git clone https://github.com/soumith/cudnn.torch.git -b R7 $ cd cudnn.torch $ luarocks make cudnn-scm-1.rockspec

conda config --add channels menpo

conda config --append channels conda-forge

conda update conda