1. An information processing apparatus comprising:
a scaling section that enlarges or reduces an input image provided to the information processing apparatus;
a scanner that uses a window of a predetermined size to sequentially scan the input image; and
a discriminator that discriminates whether the image received from the scanner is an object of interest, the discriminator comprising:
a higher node having a plurality of weak discriminators that have learned a first learning sample of a first label and a second learning sample of a second label that has a predetermined coordinate relationship with the first learning sample of the first label;
a first lower node having a plurality of weak discriminators that have learned the first learning sample of the first label based on the discrimination result of the higher node; and
a second lower node that has no weak discriminator,
wherein the first lower node performs an identification process on the input image in order to recognize an image of the first label based on the discrimination result of the higher node, and
the second lower node performs an identification process on the input image in order to identify an image of the second label based on the discrimination result of the higher node by utilizing the weak discriminators in the first lower node according to the coordinate relationship.
2. The information processing apparatus according to claim 1, wherein the weak discriminators in the higher node and the first lower node perform the identification process based on how much the difference between the values of two pixels in the image data of the learning sample differs from a threshold value, and
the second lower node performs the identification process based on how much the difference between the values of two pixels on the image according to the input signal at the positions obtained by performing coordinate conversion on the positions of the two pixels used in the identification process on the first node based on the coordinate relationship differs from the threshold value used in the identification process on the first node.
3. The information processing apparatus according to claim 2, wherein the learning sample of the second label corresponds to the image obtained by rotating the learning sample of the first label by a predetermined angle in the rolling direction or flipping the learning sample of the first label in the horizontal direction, and
the second lower node performs the identification process based on how much the difference between the values of two pixels on the image according to the input signal at the positions obtained by rotating the positions of the two pixels used in the identification process on the first node by the predetermined angle in the rolling direction or flipping the positions of the two pixels in the horizontal direction differs from the threshold value used in the identification process on the first node.
4. The information processing apparatus according to claim 3, further comprising output means that acquires a signal that undergoes the identification process and outputs it to the higher node, the first lower node or the second lower node,
wherein the output means outputs an image signal of an image larger than the learning sample to the second lower node when the second lower node performs the identification process based on how much the difference between the values of two pixels on the image according to the input signal at the positions obtained by rotating the positions of the two pixels used in the identification process on the first node by the predetermined angle in the rolling direction differs from the threshold value used in the identification process on the first node.
5. The information processing apparatus according to claim 4, wherein the weak discriminators that form the higher node and the first lower node rotate the learning sample by the predetermined angle in the rolling direction with its center aligned with the center of the image according to the image signal outputted from the output means and do not perform the learning on the areas that go off the image according to the image signal.
6. The information processing apparatus according to claim 1, wherein the higher node and the first and second lower nodes include abort means for weighting each discrimination result with reliability, accumulating the weighted discrimination results and aborting the identification process based on the accumulated sum.
7. The information processing apparatus according to claim 6, wherein the abort means judges whether or not to abort the identification process whenever the discrimination result is weighted and accumulated or whenever a predetermined number of discrimination results are weighted and accumulated.
8. A method for processing information using an information processing apparatus, wherein the information processing apparatus includes
a higher node having a plurality of weak discriminators that have learned a first learning sample of a first label and a second learning sample of a second label that has a predetermined coordinate relationship with the first learning sample of the first label,
a first lower node having a plurality of weak discriminators that have learned the first learning sample of the first label based on the discrimination result of the higher node, and
a second lower node that has no weak discriminator,
the method for processing information comprising the steps of:
using the first lower node to perform an identification process on an input signal in order to recognize an image of the first label based on the discrimination result of the higher node, and
using the second lower node to perform an identification process on the input signal in order to identify an image of the second label based on the discrimination result of the higher node by utilizing the weak discriminators in the first lower node according to the coordinate relationship.
9. A non-transitory computer readable medium on which a computer readable program is recorded for an information processing apparatus that includes
a higher node having a plurality of weak discriminators that have learned a first learning sample of a first label and a second learning sample of a second label that has a predetermined coordinate relationship with the first learning sample of the first label,
a first lower node having a plurality of weak discriminators that have learned the first learning sample of the first label based on the discrimination result of the higher node, and
a second lower node that has no weak discriminator,
the program comprising the steps of:
using the first lower node to perform an identification process on an input signal in order to recognize an image of the first label based on the discrimination result of the higher node, and
using the second lower node to perform an identification process on the input signal in order to identify an image of the second label based on the discrimination result of the higher node by utilizing the weak discriminators in the first lower node according to the coordinate relationship.
10. A non-transitory computer readable storage medium storing a program for an information processing apparatus that includes
a higher node having a plurality of weak discriminators that have learned a first learning sample of a first label and a second learning sample of a second label that has a predetermined coordinate relationship with the first learning sample of the first label,
a first lower node having a plurality of weak discriminators that have learned the first learning sample of the first label based on the discrimination result of the higher node, and
a second lower node that has no weak discriminator,
the program causing a computer to execute the steps of:
using the first lower node to perform an identification process on an input signal in order to recognize an image of the first label based on the discrimination result of the higher node, and
using the second lower node to perform an identification process on the input signal in order to identify an image of the second label based on the discrimination result of the higher node by utilizing the weak discriminators in the first lower node according to the coordinate relationship.
The claims below are in addition to those above.
All refrences to claim(s) which appear below refer to the numbering after this setence.
1-6. (canceled)
7. A method of making a QW layer for a semiconductor laser comprising depositing a layer of Ga, In, As, and Sb onto a GaAs substrate, wherein said In is included at an atomic ratio of 30% or more relative to Group-III elements, and wherein said depositing is performed with a partial pressure of Sb that is sufficient to form an active layer of about 0.3% to about 0.8% in atomic ratio of Sb relative to Group-V elements.
8. A method of making a QW layer for a semiconductor laser comprising depositing a layer of Ga, In, N, As, and Sb onto a GaAs substrate, wherein said In is included at an atomic ratio of 30% or more relative to Group-III elements, and wherein said depositing is performed with a partial pressure of Sb that is sufficient to form an active layer of about 0.2% to about 6% in atomic ratio of Sb relative to Group-V elements.
9. The method of claim 8, wherein said active layer is deposited between barrier layers of GaNzAs1-z.
10. The method of claim 9, additionally comprising heat treating said active layer after deposition at a temperature of about 675 to 725 degrees C.
11-23. (canceled)
24. A method of making a semiconductor laser comprising:
depositing a first barrier layer of GaNzAs1-z onto a substrate;
depositing an active layer of GaxIn1-xAs1-y1-y2Ny1Sby2 over said first barrier layer; and depositing a second barrier layer of GaNzAs1-z over said active layer.
25. The method of claim 24, wherein said substrate comprises GaAs.
26. The semiconductor laser of claim 24, wherein said semiconductor laser comprises a VCSEL.
27. The semiconductor laser of claim 24, wherein said semiconductor laser comprises an edge emitting laser.
28. The method of claim 24, wherein 0.31-x, wherein 0\u2266y1\u22660.03, and wherein 0.002\u2266y2\u22660.06.
29. The method of claim 24 wherein 0<z\u22660.05.
30. The method of claim 24, additionally comprising heat treating said layers at a temperature of about 675 to about 725 degrees C.
31. A method for manufacturing a semiconductor laser device comprising:
forming a laser structure by depositing a QW active layer structure over a substrate, wherein said QW active layer structure includes at least one QW layer comprising GaxIn1-xAs1-y1-y2Ny1Sby2, wherein 0.3\u22661-x, wherein 0\u2266y1\u22660.007, and wherein 0.002\u2266y2\u22660.06, and
heat treating said laser structure, after growth of said QW active layer structure at a temperature of about 570 to 630 degrees C.
32. A method for manufacturing a semiconductor laser device comprising:
forming a laser structure by depositing a QW active layer structure over a substrate, wherein said QW active layer structure includes at least one QW layer comprising GaxIn1-xAs1-y1-y2Ny1Sby2, wherein 0.31-x, wherein 0.007\u2266y1\u22660.03, and wherein 0.002\u2266y2\u22660.06, and
heat treating said laser structure, after growth of said QW active layer structure at a temperature of about 670 to 730 degrees C.